ES2417153T5 - Unstructured recombinant polymers and uses thereof - Google Patents

Description

DESCRIPTION

Unstructured recombinant polymers and uses thereof

CROSS REFERENCE BACKGROUND OF THE INVENTION

[0001]It has been well documented that protein properties, in particular plasma clearance and immunogenicity, can be improved by attaching hydrophilic polymers to these proteins (Kochendoerfer, G. (2003) Expert Opin Biol Ther, 3: 1253-61), ( Greenwald, R. B. et al. (2003) Adv Drug Deliv Rev, 55: 217-50), (Harris, J. M. et al. (2003) Nat Rev Drug Discov, 2: 214-21). Examples of polymer-modified proteins that have been cleared by the FDA for the treatment of patients are Adagen, Oncaspar, PEG-Intron, Pegasys, Somavert, and Neulasta. Many more polymer-modified proteins are in clinical trials. These polymers exert their effect by increasing the hydrodynamic radius (also called Stoke radius) of the modified protein with respect to the unmodified protein, which reduces the rate of elimination by filtration from the kidney (Yang, K. et al. (2003) Protein Eng, 16: 761-70). Additionally, polymer binding may reduce the interaction of the modified protein with other proteins, cells, or surfaces. In particular, polymer binding may reduce interactions between the modified protein and antibodies and other components of the immune system, thereby reducing the formation of a host immune response to the modified protein. Of particular interest is the modification of proteins by PEGylation, that is, joining linear or branched polyethylene glycol polymers. Reduced immunogenicity was shown after PEGylation, for example, for phenylalanine ammonia lyase (Gamez, A. et al. (2005) Mol Ther, 11: 986-9), antibodies (Deckert, P. M. et al. (2000) Int J Cancer , 87: 382-90.), staphylokinase (Collen, D. et al. (2000) Circulation, 102: 1766-72) and hemoglobin (Jin, C. et al. (2004) Protein Pept Lett, 11: 353- 60). Typically, such polymers are conjugated to the protein of interest via a chemical modification step after the unmodified protein has been purified.

[0002]Various polymers can bind to proteins. Of particular interest are hydrophilic polymers that have flexible conformations and hydrate well in aqueous solutions. A frequently used polymer is polyethylene glycol (PEG). These polymers tend to have large hydrodynamic radii with respect to their molecular weight (Kubetzko, S. et al. (2005) Mol Pharmacol, 68: 1439-54). Bound polymers tend to have limited interactions with the protein to which they are bound and thus the polymer-modified protein retains its relevant functions.

[0003]The chemical conjugation of polymers with proteins requires complex multi-step procedures. Typically, the protein component needs to be produced and purified before the chemical conjugation step. The conjugation step can result in the formation of product mixtures that need to be separated, leading to significant product loss. Alternatively, such mixtures can be used as a final pharmaceutical product. Some examples are currently marketed PEGylated interferon-alpha products used as mixtures (Wang, B. L. et al. (1998) J Submicrosc Cytol Pathol, 30: 503-9; Dhalluin, C. et al. (2005) Bioconjug Chem, 16: 504-17). Such mixtures are difficult to manufacture and characterize and contain isomers with reduced or non-therapeutic activity.

[0004]Procedures have been described that allow site-specific addition of polymers such as PEG. Examples are selective PEGylation at a single glycosylation site of the target protein or selective PEGylation of a non-natural amino acid that has been manipulated in the target proteins. In some cases it has been possible to selectively PEGylate the N terminus of a protein, while preventing PEGylation of lysine side chains in the target protein by carefully controlling the reaction conditions. Still another approach to site-specific PEGylation of target proteins is the introduction of cysteine residues that allow selective conjugation. All of these procedures have significant limitations. Selective N-terminal PEGylation requires careful control of the procedure and side reactions are difficult to eliminate. The introduction of cysteines for PEGylation can interfere with protein production and/or purification. The specific introduction of unnatural amino acids requires specific host organisms for protein production. Another limitation of PEGylation is that PEG is typically manufactured as a mixture of polymers with similar, but not uniform, length. The same limitations are inherent in many other chemical polymers.

[0005]Chemical conjugation using multifunctional polymers that would allow the synthesis of products with multi-protein modules is even more complex than the conjugation of polymers of a single protein domain.

[0006]It has recently been observed that some proteins from pathogenic organisms contain repetitive peptide sequences that appear to lead to a relatively long serum half-life of proteins containing these sequences (Alvarez, P. et al. (2004) J Biol Chem , 279: 3375-81). It has been shown that oligomeric sequences that are based on such pathogen-derived repetitive sequences can fuse with other proteins producing high serum half-life. However, these pathogen-derived oligomers have several deficiencies. Pathogen-derived sequences tend to be immunogenic. It has been described that the sequences can be modified to reduce their immunogenicity. However, attempts to remove T cell epitopes from sequences that contribute to the formation of immune reactions have not been reported. Furthermore, pathogen-derived sequences have not been optimized for pharmacological applications that require sequences with good solubility and very low affinity for other target proteins.

[0007]Thus, there is a significant need for compositions and procedures that would allow multi-polymer modules and multi-protein modules to be combined into defined multi-domain products.

SUMMARY OF THE INVENTION

[0008]The present disclosure describes an unstructured recombinant polymer (URP) comprising at least 40 contiguous amino acids, wherein said URP is substantially incapable of binding non-specifically to a whey protein, and wherein (a) the sum of the glycine (G), aspartate (D), alanine (A), serine (S), threonine (T), glutamate (E) and proline (P) residues contained in the URP constitutes more than approximately 80% of the total amino acids of the URP; and/or (b) at least 50% of the amino acids lack secondary structure as determined by the Chou-Fasman algorithm. In a related embodiment, an unstructured recombinant polymer (URP) comprising at least 40 contiguous amino acids is described, wherein said URP has a degradation half-life in serum in vivo greater than approximately 24 hours, and wherein (a) the sum of The residues of glycine (G), aspartate (D), alanine (A), serine (S), threonine (T), glutamate (E) and proline (P) contained in the URP constitute more than approximately 80% of the total URP amino acids; and/or (b) at least 50% of the amino acids lack secondary structure as determined by the Chou-Fasman algorithm. The subject URP may comprise a non-natural amino acid sequence. If desired, the URP is selected for incorporation into a heterologous protein, and wherein upon incorporation of the URP into a heterologous protein, said heterologous protein exhibits a prolonged serum secretion half-life and/or greater solubility relative to the corresponding protein that is deficient in said URP. The half-life can be extended two times, three times, five times, ten times or more. In some aspects, incorporation of the URP into a heterologous protein produces at least a 2-fold, 3-fold, 4-fold, 5-fold, or more increase in the apparent molecular weight of the protein as approximated by size exclusion chromatography. In some aspects, URPs have a T epitope score less than -3.5 (e.g., -4 or less, -5 or less). In some aspects, URPs may contain predominantly hydrophilic residues. If desired, at least 50% of the URP amino acids lack secondary structure as determined by the Chou-Fasman algorithm. The glycine residues contained in the URP may constitute at least about 50% of the total amino acids of the URP. In some aspect, any one type of single amino acids selected from the group consisting of glycine (G), aspartate (D), alanine (A), serine (S), threonine (T), glutamate (E) and proline (P ) contained in the URP constitutes more than approximately 20%, 30%, 40%, 50%, 60% or more of the total amino acids of the URP. In some aspects, the URP comprises more than about 100, 150, 200 or more contiguous amino acids.

[0009] Also described herein is a protein comprising one or more of the subject URPs, wherein the subject URPs are heterologous with respect to the protein. The total length of the aggregating URPs may exceed approximately 40, 50, 60, 100, 150, 200 or more amino acids. The protein may comprise one or more functional modules selected from the group consisting of effector module, binding module, N-terminal module, C-terminal module and any combination thereof. If desired, the subject protein comprises a plurality of binding modules, wherein the individual binding modules exhibit binding specificities for the same or different targets. The binding module may comprise a disulfide-containing scaffold formed by pairing between cysteine scaffolds. The binding module can bind to a target molecule, which target is selected from the group consisting of cell surface protein, secretase protein, cytosolic protein and nuclear protein. The target may be an ion channel and/or GPCR. If desired, the effector module can be a toxin. The protein containing the URP typically has a prolonged serum secretion half-life of at least 2, 3, 4, 5, 10 or more times relative to a corresponding protein that is deficient in said URP.

[0010] Also described herein is a non-naturally occurring protein comprising at least 3 repeat units of amino acid sequences, each of the repeat units comprising at least 6 amino acids, in which the majority of the Segments comprising about 6 to about 15 contiguous amino acids of the at least 3 repeat units are present in one or more native human proteins. In one aspect, the majority of the segments, or each segment comprising about 9 to about 15 contiguous amino acids within the repeat units, are present in one or more native human proteins. The segments may comprise about 9 to about 15 amino acids. The three repeat units may share substantial sequence homology, for example, sharing sequence identity greater than about 50%, 60%, 70%, 80%, 90%, or 100% when aligned. Such a non-natural protein may also comprise one or more modules selected from the group consisting of binding modules, effector modules, multimerization modules, C-terminal modules and N-terminal modules. If desired, the non-natural protein may comprise units of individual repeaters having the subject unstructured recombinant polymer (URP).

[0011] Also described herein are recombinant polynucleotides comprising coding sequences encoding the target URPs, URP-containing proteins, microproteins and toxins. Also provided are vectors containing the subject polynucleotides, host cells harboring the vectors, genetic packages displaying the subject URPs, URP-containing proteins, toxins, and any other proteinaceous entities disclosed herein. Also described herein are expression vector selection libraries of the present invention.

[0012]A method for producing a protein comprising an unstructured recombinant polymer (URP) is also described herein. The method involves (i) providing a host cell comprising a recombinant polynucleotide encoding the protein, said protein comprising one or more URPs, said URP comprising at least 40 contiguous amino acids, wherein said URP is substantially incapable of binding non-specifically to a whey protein, and in which (a) the sum of the residues of glycine (G), aspartate (D), alanine (A), serine (S), threonine (T), glutamate (E) and proline ( P) contained in the URP constitutes more than approximately 80% of the total amino acids of the URP; and/or (b) at least 50% of the amino acids lack secondary structure as determined by the Chou-Fasman algorithm; and (ii) cultivating said host cell in a suitable culture medium under conditions to effect expression of said protein of said polynucleotide. Suitable host cells are eukaryotic (for example, CHO cells) and prokaryotic cells.

[0013]The present invention provides a method of increasing the serum half-life of a protein comprising: fusing said protein with one or more unstructured recombinant polymers (URP), wherein the URP comprises at least about 200 contiguous amino acids, and in which (a) the sum of the residues of glycine (G), aspartate (D), alanine (A), serine (S), threonine (T), glutamate (E) and proline (P) contained in the URP constitutes more than approximately 80% of the total amino acids of URP; and (b) at least 50% of the amino acids are not present in the secondary structure as determined by the Chou-Fasman algorithm; and wherein said URP is substantially incapable of binding non-specifically to a whey protein. The URP has a T epitope score equal to or less than -4.

[0014]This document also describes a procedure for detecting the presence or absence of a specific interaction between a target protein and an exogenous one that is displayed on a genetic package, in which said protein comprises one or more non-recombinant polymers. structured (URP), the method comprising: (a) providing a genetic package that displays a protein comprising one or more unstructured recombinant polymers (URP); (b) contacting the genetic package with the target under suitable conditions to produce a stable protein-target complex; and (c) detect the formation of the stable protein-target complex on the genetic package, thus detecting the presence of a specific interaction. The method may further comprise obtaining a nucleotide sequence of the genetic package that encodes the exogenous protein. In some aspects, the presence or absence of a specific interaction is between the URP and a target comprising a serum protein. In some aspects, the presence or absence of a specific interaction is between the URP and a target comprising a serum protease.

[0015]This document also describes a genetic package that displays a microprotein, in which said microprotein retains the binding capacity to its native target. In some aspects, the microprotein exhibits binding capacity toward at least one family of ion channels selected from the group consisting of a sodium channel, a calcium potassium channel, an acetylcholine channel, and a chloride channel. If desired, the microprotein is an ion channel binding microprotein, and is modified such that (a) the microprotein binds to a different channel family relative to the corresponding unmodified microprotein; (b) the microprotein binds to a different subfamily of the same channel family relative to the corresponding unmodified microprotein; (c) the microprotein binds to a different species of the same channel subfamily relative to the corresponding unmodified microprotein; (d) the microprotein binds to a different site on the same channel relative to the corresponding unmodified microprotein; and/or (e) the microprotein binds to the same site of the same channel but gives a different biological effect with respect to the corresponding unmodified microprotein. In some respects, the microprotein is a toxin. A library of genetic packages that displays the target microproteins and/or toxins is also described. If desired, the genetic package envisions a proteinaceous toxin that retains some or all of its toxicity spectrum. The toxin may be derived from a single toxin protein, or derived from a family of toxins. Also described herein is a gene package library in which the library displays a family of toxins, where the family retains some or all of its native toxicity spectrum.

[0016] Also described herein is a protein comprising a plurality of ion channel binding domains, wherein individual domains are microprotein domains that have been modified such that (a) the microprotein domains bind to a different channel family with respect to the corresponding unmodified microprotein domains; (b) the microprotein domains bind to a different subfamily of the same channel family relative to the corresponding unmodified microprotein domains; (c) the microprotein domains bind to a different species of the same subfamily relative to the corresponding unmodified microprotein domains; (d) the microprotein domains bind to a different site on the same channel relative to the corresponding unmodified microprotein domains; (e) the microprotein domains bind to the same site of the same channel, but give a different biological effect relative to the corresponding unmodified microprotein domains; and/or (f) the microprotein domains bind to the same site of the same channel and give the same biological effect with respect to the corresponding unmodified microprotein domains.

[0017]This document also describes a procedure for obtaining a microprotein with the desired property, which comprises: (a) providing an object library; and (b) screening the selection library to obtain at least one phage display of a microprotein with the desired property. Also provided are polynucleotides, vectors, genetic packages, host cells for use in any one of the disclosed methods.

BRIEF DESCRIPTION OF THE DRAWINGS

[0018]The novel features of the invention are set out with particularity in the attached claims. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description which sets forth illustrative embodiments, in which the principles of the invention are used, and the accompanying drawings of which:

[0019]FIG. 1 shows the modular components of a MURP. Binding modules, effector modules and multimerization modules are represented as circles. URP modules, N-end and C-end modules are shown as rectangles.

[0020]FIG. 2 shows examples of modular MURP architectures. Binding modules (BM) in a MURP can have identical or different target specificities.

[0021]FIG. 3 shows that a repeat protein that is based on a human sequence may contain novel amino acid sequences, which may contain T cell epitopes. These novel sequences are formed at the splicing between neighboring repeat units.

[0022]FIG. 4 illustrates the design of a URP sequence that is a repeat protein based on three human donor sequences D1, D2, and D3. The repeat unit of this URP was chosen so that even 9-mer sequences that span the splice between neighboring units could be found in at least one of the human donor sequences.

[0023]FIG. 5 Example of a URP sequence that is a repeat protein based on the sequences of three human proteins. The lower portion of the figure illustrates that all 9-mer subsequences in the URP occur in at least one of the human donor proteins.

[0024] FIG. 6 Example based on URP sequence based on human POU domain residues 146 182.

[0025] FIG. 7 shows the advantage of separating modules with information-rich sequences by inserting URP modules between such sequences. The left side of the figure shows that direct fusion of modules A and B leads to novel sequences in the splicing region. These splicing sequences may be epitopes. The right half of the figure shows that the insertion of a URP module between module A and B prevents the formation of such splice sequences containing partial sequences of modules A and B. Instead, the ends of modules A and B give splicing sequences containing<u>R<p>sequences and are thus predicted to have low immunogenicity.

[0026] FIG. 8 shows drug delivery constructs that are based on URP. Drug molecules represented as hexagons are chemically conjugated to MURP.

[0027] FIG. 9 sample and MURP containing a protease-sensitive site. The URP module is designed in such a way that it blocks the function of the effector module. Protease cleavage eliminates a part of the URP module and produces high effector function activity.

[0028] FIG. 10 shows how a URP module can act as a linker between a binding module and an effector module. The binding module can bind to a target and as a consequence the local concentration of effector module in the vicinity of the target increases.

[0029] FIG. 11 shows a procedure for constructing genes encoding URP sequences from libraries of short URP modules. The URP module library can be inserted into a filler vector containing green fluorescent protein (GFP) as a reporter to facilitate the identification of highly expressed URP sequences. The figure illustrates that genes encoding long URP sequences can be constructed by iterative dimerization.

[0030] FIG. 12 shows MURPs containing multiple binding modules for death receptors. Death receptors are triggered by trimerization and thus MURPs contain at least three binding elements for a death receptor particularly potent in inducing cell death. The lower portion of the figure illustrates that the specificity of MURP for diseased tissue can be increased by adding one or more binding modules with specificity for tumor tissue.

[0031] FIG. 13 shows a MURP comprising four binding modules (rectangles) with specificity for a tumor antigen with an effector module such as interleukin 2.

[0032] FIG. 14 shows the flowchart for the construction of URP modules with 288 residues. URP modules were constructed as fusion proteins with GFP. Libraries of URP modules with 36 amino acids were constructed first, followed by iterative dimerization to give URP modules with 288 amino acids (rPEG_H288 and rPEG_J288).

[0033] FIG. 15 Amino acid and nucleotide sequence of a URP module with 288 amino acids (rPEG_J288).

[0034] FIG. 16 Amino acid and nucleotide sequence of a URP module with 288 amino acids (rPEG_H288).

[0035] FIG. 17 Amino acid sequence of a serine-rich sequence region of the human dentin sialophosphoprotein protein.

[0036] FIG. 18 shows a sustained release derivative of a MURP. The protein contains two cysteine residues that can form a weak SS bridge. The protein can be made with the SS bridge intact. It can be formulated and injected into patients in reduced form. After injection it will oxidize in proximity to the injection site and as a result may form a high molecular weight polymer with very limited diffusivity. Active MURP can slowly leak from the injection site by limited proteolysis or limited reduction of the cross-linking SS bond.

[0037] FIG. 19 shows a sustained release form of a MURP. MURP has very limited diffusivity at the injection site and can be released from the injection site by limited proteolysis.

[0038] FIG. 20 shows a sustained release form of a MURP containing a histidine-rich sequence. MURP can be formulated and injected in combination with insoluble beads containing immobilized nickel. MURP binds to nickel beads at the injection site and is slowly released into the circulation.

[0039] FIG. 21 shows MURPs containing multimerization modules. The top part of the figure shows a MURP containing a dimerization sequence. As a result, it forms a dimer that effectively doubles its molecular weight. The center of the figure shows three MURP designs comprising two multimerization sequences. Such MURPs can form multimers with very high effective molecular weight. The bottom part of the figure illustrated a MURP containing multiple RGD sequences that are known to bind to cell surface receptors and thus confer half-life.

[0040] FIG. 22 shows a variety of MURPs that are designed to block or modulate the function of ion channels. Circles indicate binding modules with ion channel specificity. These binding modules may be derived from or identical to natural toxins with affinity for ion channel receptors. The figure illustrates that other binding domains can be added to either side of the specific ion channel binding modules, thus conferring MURPs with high efficacy or specificity for a particular cell type.

[0041] FIG. 23 shows various MURP designs for high half-life. High effective molecular weight can be achieved by increasing chain length (A), chemical multimerization (B), adding multiple copies of binding modules to a molecule separated by non-binding sites (C), construction of C-like chemical multimers ( D, E), which include multimerization sequences (F).

[0042] FIG. 24 shows MURPs that can be formed by chemical conjugation of binding modules with a recombinant URP sequence. The URP sequence is designed to contain multiple lysine (K) residues as conjugation sites.

[0043] FIG. 25 shows the design of a 2SS binding module library. The sequences contain a constant 1SS sequence in the center that is flanked by random sequences containing cysteine residues at varying distances from the 1Ss core.

[0044] FIG. 26 shows the design of a 2SS binding module library. The sequences contain a constant 1SS sequence in the center that is flanked by random sequences containing cysteine residues at varying distances from the 1S<s>core.

[0045] FIG. 27 shows the design of a 1SS binding module dimer library. Initially, a collection of 1SS binding modules is amplified by two PCR reactions. The resulting PCR products are combined and generate dimers in a subsequent PCR step.

[0046] FIG. 28 shows Western analyzes of a fusion protein containing the 288 amino acid URP sequence rPEG_J288 after incubation for up to 3 days in 50% mouse serum.

[0047] FIG. 29 shows the results of a binding assay test for pre-existing antibodies against a 288 amino acid URP sequence.

[0048] FIG. 30 shows the binding of MURPs containing one (monomer), two (dimer), four (tetramer) or zero (rPEG36) binding modules with specificity for VEGF that were coated to microtiter plates.

[0049] FIG. 31 shows the amino acid sequence of a MURP with specificity for EpCAM. The sequence contains four binding modules with affinity for EpCAM (underlined). The sequence contains an N-terminal Flag sequence that contains the only two lysine residues in the entire sequence.

[0050] FIG. 32 shows the design of 1SS addition libraries. Random 1SS modules can be added to the N or C end of a preselected junction module or simultaneously to both sides.

[0051] FIG. 33 shows the alignment of sequences related to three-finger toxin. The figure also shows a 3D structure that was solved by NMR.

[0052]FIG. 34 shows the design of a library based on the three-finger toxin. The residues designated X were randomized. The codon choice for each random position is indicated.

[0053]FIG. 35 shows the alignment of plexin-related sequences.

[0054]FIG. 36 shows the design of a plexin-based library. The residues designated X were randomized. The codon choice for each random position is indicated.

[0055]FIG. 37 Sequences of plexin-related binding modules with specificity for DR4, ErbB2 and HGFR.

[0056]FIG. 38 shows a binding assay for microprotein-based binding domains with specificity for VEGF.

[0057]FIG. 39 shows sequences of 2SS and 3SS binding modules that were isolated from manufacturing libraries with VEGF specificity. The top part of the protein shows PAGE gel analysis of the proteins purified by heat-lysis.

[0058]FIG. 40 shows cloning steps to construct the rPEG_J72 URP sequence.

[0059]FIG. 41 shows the construction of a library of URP modules with 36 amino acids called rPEG_J36. The region encoding rPEG_J36 was assembled by ligating three shorter segments encoding rPEG_J12 and a stopper module.

[0060]FIG. 42 shows the nucleotide sequence and translation of the stuffing vector pCW0051. The filler region is flanked by BsaI and BbsI sites and contains multiple stop codons.

[0061]FIG. 43 shows a PAGE gel of the purification of the rPEG_J288 URP fused to GFP. Lane 2 shows cell lysate; lane 3: product purified by IMAC; lane 4: product purified by anti-Flag.

[0062]FIG. 44 Amino acid sequence of fusion proteins between rPEG_J288 and the human interferon alpha, G-CSF and human growth hormone effector domains.

[0063]FIG. 45 shows the Western analysis of expression of fusion proteins between rPEG_J288 and human growth hormone (lanes 1 and 2), alpha interferon (lanes 3 and 4) and GFP (lanes 5 and 6). Both soluble and insoluble material was analyzed for each protein.

[0064]FIG. 46 shows the design of MURP based on the OSK1 toxin. The figure shows which URP sequences and/or binding modules can be added to either side of OSK1

[0065]FIG. 47 represents, by way of example, product formats that comprise the object URPs. DETAILED DESCRIPTION OF THE INVENTION

[0066]Although preferred embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are only provided by way of example. Numerous variations, changes and substitutions will now occur to those skilled in the art without departing from the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of the invention. The following claims are intended to define the scope of the invention and thus cover procedures and structures within the scope of these claims and their equivalents.

General techniques:

[0067]The practice of the present invention employs, unless otherwise indicated, conventional techniques of immunology, biochemistry, chemistry, molecular biology, microbiology, cell biology, genomics and recombinant DNA, which are within the skill in the art. . See Sambrook, Fritsch and Maniatis, MOLECULAR CLONING: A LABORATORY MANUAL, 2nd edition (1989); CURRENT PROTOCOLS IN MOLECULAR BIOLOGY (F. M. Ausubel et al. eds., (1987)); the METHODS IN ENZYMOLOGY series (Academic Press, Inc.): PCR 2: A PRACTICAL APPROACH (M.J. MacPherson, B.D. Hames and G.R. Taylor eds. (1995)), Harlow and Lane, eds. (1988) ANTIBODIES, A LABORATORY MANUAL, and ANIMAL CELL CULTURE (R.I. Freshney, ed. (1987)).

Definitions:

[0068]As used in the specification and claims, the singular form “a”, “an”, “the” and “the” include plural references, unless the context clearly dictates otherwise. For example, the term "a cell" includes a plurality of cells, including mixtures thereof.

[0069]The terms “polypeptide”, “peptide”, “amino acid sequence” and “protein” are used interchangeably herein to refer to polymers of amino acids of any length. The polymer may be linear or branched, may comprise modified amino acids, and may be interrupted by non-amino acids. The terms also encompass a polymer of amino acids that has been modified, for example, disulfide bond formation, glycosylation, lipidation, acetylation, phosphorylation, or any other manipulation such as conjugation with a brand component. As used herein, the term “amino acid” refers to both natural and/or non-natural and synthetic amino acids including, but not limited to, glycine and both the D and L optical isomers, and amino acid analogs. and peptidomimetics. Single-letter or conventional three-letter codes are used to designate amino acids.

[0070]A “repetitive sequence” refers to an amino acid sequence that can be described as an oligomer of repeating peptide sequences, which form direct repeats, or inverse repeats or alternating repeats of multiple sequence motifs. These repeat oligomer sequences may be identical or homologous to each other, but may also be multiple repeat motifs. Repetitive sequences are characterized by a very low information content. A repetitive sequence is not a required feature of a URP and in some cases a non-repetitive sequence will in fact be preferred.

[0071]Amino acids can be characterized based on their hydrophobicity. Several scales have been developed. An example is a scale developed by Levitt, M et al. (see Levitt, M (1976) J Mol Biol 104, 59, no. 3233, listed in Hopp, TP et al. (1981) Proc Natl Acad Sci U S A 78, 3824, no. 3232). Examples of "hydrophilic amino acids" are arginine, lysine, threonine, alanine, asparagine and glutamine. Of particular interest are the hydrophilic amino acids aspartate, glutamate and serine, and glycine. Examples of “hydrophobic amino acids” are tryptophan, tyrosine, phenylalanine, methionine, leucine, isoleucine and valine.

[0072]The term “denatured conformation” describes the state of a peptide in solution that is characterized by great conformational freedom of the peptide backbone. Most peptides and proteins adopt a denatured conformation in the presence of high concentrations of denaturants or at elevated temperatures. Peptides in denatured conformation have characteristic CD spectra and are generally characterized by a lack of long-range interactions as determined by, for example, NMR. Denatured conformation and unfolded conformation will be used synonymously.

[0073]The terms “unstructured protein sequences (UNP)” and unstructured recombinant polymer” (URP) are used interchangeably herein. The terms refer to amino acid sequences that share common characteristics with denatured peptide sequences, for example, exhibiting typical behavior similar to denatured peptide sequences, under physiological conditions, as detailed herein. URP sequences lack a defined tertiary structure and have limited or no secondary structure as detected, for example, by the Chou-Fasman algorithm.

[0074]As used herein, the term “cell surface proteins” refers to components of the plasma membrane of a cell. It encompasses integral and peripheral membrane proteins, glycoproteins, polysaccharides and lipids that constitute the plasma membrane. An integral membrane protein is a transmembrane protein that spans the lipid bilayer of a cell's plasma membrane. A typical integral membrane protein consists of at least one membrane-spanning segment that generally comprises hydrophobic amino acid residues. Peripheral membrane proteins do not extend into the hydrophobic interior of the lipid bilayer and are attached to the membrane surface by covalent or noncovalent interaction directly or indirectly with other membrane components.

[0075]The terms “membrane”, “cytosolic”, “nuclear” and “secreted” as applied to cellular proteins specify the extracellular and/or subcellular location in which the cellular protein is primarily, predominantly or preferentially located.

[0076] “Cell surface receptors” represent a subset of membrane proteins, which can bind their respective ligands. Cell surface receptors are molecules anchored on or inserted into the plasma membrane of the cell. They constitute a large family of proteins, glycoproteins, polysaccharides and lipids, which serve not only as structural constituents of the plasma membrane, but also as regulatory elements that govern a variety of biological functions.

[0077]The term “module” refers to a part of a protein that is physically or functionally differentiated from other portions of the protein or peptide. A module may comprise one or more domains. In general, a module or domain can be a single stable three-dimensional structure, regardless of size. The tertiary structure of a typical domain is stable in solution and remains the same whether such a member is isolated or covalently fused to other domains. A domain generally has a particular tertiary structure formed by spatial relationships of secondary structure elements, such as beta sheets, alpha helices, and unstructured loops. In domains of the microprotein family, disulfide bridges are generally the primary elements that determine the tertiary structure. In some cases, domains are modules that can confer specific functional activity, such as avidity (multiple binding sites for the same target), multispecificity (binding sites for different targets), half-life (using a domain, cyclic peptide or peptide linear) that binds to a serum protein such as human serum albumin (HSA) or to IgG (hIgG1, 2, 3 or 4) or to red blood cells. Functionally defined domains have distinct biological function(s). The ligand binding domain of a receptor is, for example, the domain that binds ligand. An antigen-binding domain refers to the part of an antigen-binding unit or an antibody that binds to the antigen. Functionally defined domains do not need to be encoded by contiguous amino acid sequences. Functionally defined domains may contain one or more physically defined domains. Receptors, for example, are generally divided into the extracellular ligand-binding domain, a transmembrane domain, and an intracellular effector domain. A “membrane anchoring domain” refers to the portion of a protein that mediates membrane association. Generally, the membrane anchoring domain is composed of hydrophobic amino acid residues. Alternatively, the membrane anchoring domain may contain modified amino acids, for example, amino acids that are linked to a fatty acid chain, which in turn anchors the protein to a membrane.

[0078]“Not naturally occurring” as applied to a protein means that the protein contains at least one amino acid that is different from the corresponding natural or native protein. Unnatural sequences can be determined by searching with BLAST using, for example, the smallest probability of minimum sum in which the comparison window is the length of the sequence of interest (the searched one) and when compared to the non-redundant database (“nr”) from Genbank using BLAST 2.0. The BLAST 2.0 algorithm, described in Altschul et al. (1990) J. Mol. Biol. 215:403-410, respectively. Software for performing BLAST analysis is publicly available from the National Center for Biotechnology Information.

[0079]A “host cell” includes an individual cell or cell culture that may be or has been a recipient for the subject vectors. Host cells include progeny of a single host cell. The progeny may not necessarily be completely identical (in morphology or in genomics of the total DNA complement) to the original parental cell due to natural, accidental or intentional mutation. A host cell includes cells transfected in vivo with a vector described herein.

[0080]As used herein, the term "isolated" means separated from constituents, cellular and otherwise, in which the polynucleotide, peptide, polypeptide, protein, antibody, or fragments thereof, are normally associated In nature. As is evident to those skilled in the art, a non-naturally occurring polynucleotide, peptide, polypeptide, protein, antibody, or fragments thereof does not require “isolation” to distinguish it from its naturally occurring counterpart. Furthermore, a “concentrated,” “separated,” or “diluted” polynucleotide, peptide, polypeptide, protein, antibody, or fragments thereof is differentiable from its naturally occurring counterpart because the concentration or number of molecules per volume is greater. than “concentrated” or lower than “separated” than its naturally occurring counterpart.

[0081]“Linked” and “fused” or “fusion” are used interchangeably herein. These terms refer to joining together two or more chemical elements or components, by any means including chemical conjugation or recombinant means. An “in-frame merge” refers to the joining of two or more open reading frames (OFRs) to form a longer continuous OFR, in a way that maintains the correct reading frame of the original OFRs. Thus, the resulting recombinant fusion protein is a single protein containing two or more segments that correspond to polypeptides encoded by the original OFRs (segments that are not normally so linked in nature.)

[0082]In the context of polypeptides, a “linear sequence” or “sequence” is an order of amino acids in a polypeptide in an amino- to carboxyl-terminal direction in which residues that are neighbors of each other in the sequence are contiguous. in the primary structure of the polypeptide. A “partial sequence” is a linear sequence of part of a polypeptide that is known to comprise additional residues in one or both directions.

[0083]“Heterologous” means derived from an entity genotypically different from the rest of the entity with which it is being compared. For example, a glycine-rich sequence removed from its native coding sequence and operatively linked to a coding sequence other than the native sequence is a heterologous glycine-rich sequence. The term "heterologous" as applied to a polynucleotide, a polypeptide, means that the polynucleotide or polypeptide is derived from an entity genotypically different from that of the rest of the entity with which it is being compared.

[0084]The terms “polynucleotides”, “nucleic acids”, “nucleotides” and “oligonucleotides” are used interchangeably. They refer to a polymeric form of nucleotides of any length, both deoxyribonucleotides and ribonucleotides, or analogues thereof. Polynucleotides can have any three-dimensional structure, and can perform any function, known or unknown. The following are non-limiting examples of polynucleotides: coding or non-coding regions of a gene or gene fragment, loci defined from linkage analysis, exons, introns, messenger RNA (mRNA), transfer RNA, ribosomal RNA, ribozymes , cDNA, recombinant polynucleotides, branched polynucleotides, plasmids, vectors, DNA isolated from any sequence, RNA isolated from any sequence, nucleic acid probes and primers. A polynucleotide may comprise modified nucleotides, such as methylated nucleotides and nucleotide analogs. If present, modifications to the nucleotide structure may be conferred before or after polymer assembly. The nucleotide sequence can be interrupted by non-nucleotide components. A polynucleotide may be further modified after polymerization, such as by conjugation with a label component.

[0085]“Recombinant” as applied to a polynucleotide means that the polynucleotide is the product of various combinations of cloning, restriction and/or ligation steps, and other procedures that produce a construct that is distinct from a polynucleotide found in nature. .

[0086]The terms “gene” or “gene fragment” are used interchangeably herein. They refer to a polynucleotide that contains at least one open reading frame that can encode a particular protein after being transcribed and translated. A gene or gene fragment may be genomic DNA or cDNA, as long as the polynucleotide contains at least one open reading frame that can cover the entire coding region or a segment thereof. A “fusion gene” is a gene composed of at least two heterologous polynucleotides that are linked together.

[0087]A “vector” is a nucleic acid molecule, preferably self-replicating, that transfers an inserted nucleic acid molecule into and/or between host cells. The term includes vectors that function primarily for the insertion of DNA or RNA into a cell, replication vectors that function primarily for the replication of DNA or RNA, and expression vectors that function for transcription and/or translation of the DNA or RNA. Vectors that provide more than one of the above functions are also included. An “expression vector” is a polynucleotide that, when introduced into an appropriate host cell, can be transcribed and translated into a polypeptide(s). An “expression system” typically connotes a suitable host cell comprising an expression vector that can function to give a desired expression product.

[0088]The “target” as used in the context of MURP is a molecule or biochemical structure with which the binding module or the URP-linked binding module can bind and in which the binding event produces a biological activity desired. The target may be a protein ligand or receptor that is inhibited, activated or otherwise acted upon by the protein. Examples of targets are hormones, cytokines, antibodies or antibody fragments, cell surface receptors, kinases, growth factors and other biochemical structures with biological activity.

[0089] A “functional module” can be any non-URP in a protein product. Thus, a functional module can be a binding module (BM), an effector module (EM), a multimerization module (MM), a C-terminal module (CM) or an N-terminal module (NM). In general, functional modules are characterized by a high information content of their amino acid sequence, that is, they contain many different amino acids and many of these amino acids are important for the function of a functional module. A functional module typically has secondary and tertiary structure, may be a folded protein domain, and may contain 1, 2, 3, 4, 5, or more disulfide bonds.

[0090] The term 'microproteins' refers to a classification in the SCOP database. Microproteins are typically the smallest proteins with a fixed structure and typically, but not exclusively, have as few as 15 amino acids with two disulfides or up to 200 amino acids with more than ten disulfides. A microprotein may contain one or more microprotein domains. Some microprotein domains or families of domains may have multiple more or less stable structures and multiple more or less similar structures that are conferred by different disulfide bond patterns, so the term stable is used in a relative way to differentiate microproteins from peptides and domains. of non-microprotein. Most microprotein toxins are composed of a single domain, but cell surface receptor microproteins frequently have multiple domains. Microproteins can be so small because their folding is stabilized by both disulfide bonds and ions such as calcium, magnesium, manganese, copper, zinc, iron, or a variety of other multivalent ions, rather than by the hydrophobic core. typical.

[0091] The term “scaffold” refers to the minimal polypeptide 'structural region' or 'sequence motif' that is used as a common conserved sequence in the construction of protein libraries. Among the fixed or conserved residues/positions of the scaffold are variable and hypervariable positions. A variety of amino acids are provided in the variable regions between the fixed scaffold residues to provide specific binding to a target molecule. A scaffold is typically defined by the conserved residues observed in an alignment of a family of sequence-related proteins. Fixed residues may be required for folding or structure, especially if the functions of the aligned proteins are different. A complete description of a microprotein scaffold may include the number, position or spacing, and binding pattern of the cysteines, as well as the position and identity of any fixed residues in the loops, including binding sites for ions such as calcium.

[0092] The “fold” of a microprotein is broadly defined by the bonding pattern of the disulfide bonds (i.e., 1-4, 2-6, 3-5). This pattern is a topological constant and is generally not flexible to conversion into another pattern without unbinding and religating the disulfides such as through reduction and oxidation (redox agents). In general, natural proteins with related sequences adopt the same disulfide bond patterns. The main determinants are the cysteine distance pattern (CDP) and some fixed non-cis residues, as well as a metal binding site, if present. In some cases, protein folding is also influenced by surrounding sequences (i.e., pro-peptides) and in some cases by chemical derivatization (e.g., gamma-carboxylation) of residues that allow the protein to bind ions. divalent metals (i.e. Ca++) that helps in their folding. For the vast majority of microproteins, no such assistance in folding is required.

[0093] However, proteins with the same binding pattern may still comprise multiple folds, based on differences in the length and composition of the loops that are large enough for the protein to give a quite different structure. An example is the conotoxin, cyclotoxin, and annatto domain families that have the same DBP, but a very different CDP, and are considered to be different folds. Determinants of a protein fold are any attribute that greatly alters the structure with respect to a different fold, such as the number and binding pattern of cysteines, the spacing of cysteines, differences in the sequence motifs of loops between cysteines (especially fixed loop residues that are likely to be needed for folding, or in the location or composition of the calcium (or other metal or co-factor) binding site.

[0094] The term “disulfide bond pattern” or “DBP” refers to the bonding pattern of cysteines, which are numbered 1-n from the N terminus to the C terminus of the protein. The disulfide bond patterns are topologically constant, which means that they can only be changed by not binding one or more disulfides such as using redox conditions. Possible patterns of 2, 3, and 4 disulfide bonds are listed below in paragraphs 0048-0075.

[0095]The term “cysteine distance pattern” or “CDP” refers to the number of non-cysteine amino acids that separate cysteines on a linear protein chain. Various notations are used: C5C0C3C equals C5CC3C equals CxxxxxCCxxxC.

[0096]The term 'Position n6' or 'n7=4' refers to the loops between cysteines and 'n6' is defined as the loop between C6 and C7; 'n7=4' means the loop between C7 and C8 is 4 amino acids long, not counting cysteines

[0097]Resistance to degradation in serum - Proteins can be removed by degradation in the blood, which usually involves proteases in serum or plasma. Resistance to degradation in serum is measured by combining the protein with human (or mouse, rat, monkey, as appropriate) serum or plasma, typically over a range of days (i.e., 0.25, 0.5, 1, 2, 4, 8, 16 days) at 37 °C. Samples for these time points are then run on a Western assay and the protein is detected with an antibody. The antibody can be a mark on the protein. If the protein shows a single band on Western, in which the size of the protein is identical to that of the injected protein, then no degradation has occurred. The time point at which 50% of the protein is degraded, as judged by Western, is the serum degradation half-life of the protein.

[0098]Whey Protein Binding - While MURP typically has several modules that bind to cell surface targets and/or serum proteins, it is desired that URP be substantially devoid of unintended activities. The URP should be designed to minimize interaction with (binding to) serum proteins, including antibodies. Different URP designs can be screened for binding to serum proteins by ELISA, by immobilizing the serum proteins and then adding the URP, incubating, washing, and then detecting the amount of bound URP. One approach is to detect the URP using an antibody that recognizes a tag that has been bound to URP. A different approach is to immobilize the URP (such as by fusion with GFP) and contact it with human serum, incubating, washing, and then detecting the amount of human antibodies still bound to the URP using secondary antibodies such as goat directed against human IgG. . Using these approaches, the present inventors have designed their URPs that show very low levels of binding to serum proteins. However, in some applications binding to serum proteins or proteins exposed to serum is desired, for example, because it may further prolong the secretion half-life. In such cases, these same assays can be used to design URPs that bind to serum proteins or serum-exposed proteins such as HSA or IgG. In other cases, MURPs may be given binding modules containing peptides that have been designed to bind serum proteins or serum-exposed proteins such as HSA or IgG.

Unstructured recombinant polymers (URP):

[0099]This specification describes the design of unstructured recombinant polymers (URP). The subject URPs are particularly useful for generating recombinant proteins of therapeutic and/or diagnostic value. Object URPs have one or more of the following characteristics.

[0100]The subject URPs comprise amino acid sequences that typically share common characteristics with peptide sequences denatured under physiological conditions. URP sequences normally behave like denatured peptide sequences under physiological conditions. URP sequences lack well-defined secondary and tertiary structures under physiological conditions. A variety of procedures have been established in the art to determine the secondary and tertiary structures of a given polypeptide. For example, the secondary structure of a polypeptide can be determined by CD spectroscopy in the “far UV” spectral region (190-250 nm). The alpha helix, beta sheet, and random coil structures each give rise to a characteristic shape and magnitude of CD spectra. Secondary structure can also be determined by certain computer programs or algorithms such as the Chou-Fasman algorithm (Chou, P. Y. et al. (1974) Biochemistry, 13: 222-45). For a given URP sequence, the algorithm can predict whether or not there is any secondary structure or absolutely no secondary structure. In general, URP sequences will have spectra that resemble denatured proteins due to their low degree of secondary and tertiary structure. If desired, URP sequences can be designed to have predominantly denatured conformations under physiological conditions. URP sequences typically have a high degree of conformational flexibility under physiological conditions and tend to have large hydrodynamic radii (Stokes radius) compared to globular proteins of similar molecular weight. As used herein, physiological conditions refers to a set of conditions including temperature, salt concentration, pH that mimic those conditions of a living subject. A multitude of physiologically relevant conditions have been established for use in in vitro assays. Generally, a physiological buffer contains a physiological concentration of salt and adjusted to a neutral pH ranging from about 6.5 to about 7.8, and preferably from about 7.0 to about 7.5. A variety of physiological tampons are listed in Sambrook et al. (1989), above, and hence is not detailed in this document. The physiologically relevant temperature ranges from about 25°C to about 38°C, and preferably from about 30°C to about 37°C.

[0101]The target URPs may be sequences with low immunogenicity. The low immunogenicity may be a direct result of the conformational flexibility of the URP sequences. Many antibodies recognize so-called conformational epitopes on protein antigens. Conformational epitopes are formed by regions of the protein surface that are composed of multiple discontinuous amino acid sequences of the protein antigen. The precise folding of the protein brings these sequences into a special well-defined configuration that can be recognized by antibodies. Preferred URPs are designed to avoid the formation of conformational epitopes. For example, of particular interest are URP sequences that have a low tendency to adapt to compactly folded conformations in aqueous solution. In particular, low immunogenicity can be achieved by choosing sequences that resist antigen processing in antigen-presenting cells, by choosing sequences that do not bind MHC well, and/or by choosing sequences that are derived from human sequences.

[0102]The target URPs may be sequences with a high degree of protease resistance. Protease resistance may also be a result of conformational flexibility of URP sequences. Protease resistance can be engineered by avoiding known protease recognition sites. Alternatively, protease-resistant sequences can be selected by phage display or related techniques from libraries of random or semi-random sequences. If desired for special applications, such as slow release of a controlled release protein preparation, serum protease cleavage sites can be constructed in a URP. Of particular interest are URP sequences with high stability (e.g., long half-life in serum, less prone to cleavage by proteases present in body fluid) in blood.

[0103]The subject URP can also be characterized by the effect whereby after incorporation into a protein, the protein has a prolonged serum half-life and/or greater solubility with respect to the corresponding protein that is deficient in URP. [Methods for determining serum half-life are known in the art (see, for example, Alvarez, P. et al. (2004) J Biol Chem, 279: 3375-81). Whether the resulting protein has a prolonged serum half-life relative to the unmodified protein can be readily determined by practicing any procedure available in the art or exemplified herein.

[0104]The subject URP may be of any length necessary to effect (a) extension of the serum half-life of a protein comprising the URP; pws. optionally (b) an increase in the solubility of the resulting protein; (c) high resistance to proteases; and/or (d) a reduced immunogenicity of the resulting protein comprising the URP. The subject URP has approximately 200, 300, 400 or more contiguous amino acids. When incorporated into a protein, the URP can be fragmented so that the resulting protein contains multiple URPs, or multiple fragments of URPs.

[0105]The URP may have an isoelectric point (pl) of 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5, 10.0, 10.5, 11.0, 11, 5, 12.0, 12.5 or even 13.0.

[0106]In general, URP sequences are rich in hydrophilic amino acids and contain a low percentage of hydrophobic or aromatic amino acids. Suitable hydrophilic residues include, but are not limited to, glycine, serine, aspartate, glutamate, lysine, arginine and threonine. Hydrophobic residues that are less favored in URP construction include tryptophan, phenylalanine, tyrosine, leucine, isoleucine, valine, and methionine. URP sequences can be rich in glycine, but URP sequences can also be rich in the amino acids glutamate, aspartate, serine, threonine, alanine or proline. Thus, the predominant amino acid can be G, E, D, S, T, A or P. The inclusion of proline residues tends to reduce sensitivity to proteolytic degradation.

[0107]The inclusion of hydrophilic residues normally increases the solubility of URP in water and aqueous media under physiological conditions. As a result of their amino acid composition, URP sequences have a low tendency to form aggregates in aqueous formulations and the fusion of URP sequences to other proteins or peptides tends to enhance their solubility and reduce their tendency to form aggregates, which is a separate mechanism to reduce immunogenicity.

[0108]URP sequences can be designed to prevent certain amino acids from conferring undesirable properties to the protein. For example, URP sequences can be designed that contain few or none of the following amino acids: cysteine (to prevent disulfide formation and oxidation), methionine (to prevent oxidation), asparagine, and glutamine (to prevent deamidation).

URP rich in glycine:

[0109]In one embodiment, the subject URP comprises a glycine-rich sequence (GRS). For example, glycine may be predominantly present such that they are the most prevalent residues present in the sequence of interest. In another example, the URP sequences can be designed so that glycine residues constitute at least about 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75 %, 80%, 85%, 90%, 95%, 100% of total amino acids. URP can also contain 100% glycines. In yet another example, URPs contain at least 30% glycine and the total concentration of tryptophan, phenylalanine, tyrosine, valine, leucine and isoleucine is then less than 20%. In another further example, URPs contain at least 40% glycine and the total concentration of tryptophan, phenylalanine, tyrosine, valine, leucine and isoleucine is then less than 10%. In yet another example, URPs contain at least about 50% glycine and the total concentration of tryptophan, phenylalanine, tyrosine, valine, leucine and isoleucine is then less than 5%.

[0110]The length of GRS can vary between about 5 amino acids and 200 amino acids or more. For example, the length of an individual contiguous GRS may contain 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 70, 80, 90,100, 120, 140, 160, 180, 200, 240, 280, 320 or 400 or more amino acids. The GRS may comprise glycine residues at both ends.

[0111]GRS may also have a significant content of other amino acids, for example, Ser, Thr, Ala or Pro. GRS may contain a significant fraction of negatively charged amino acids including, but not limited to, Asp and Glu. The GRS may contain a significant fraction of positively charged amino acids including, but not limited to, Arg or Lys. If desired, URPs can be designed to contain only a single type of amino acid (i.e., Gly or Glu), sometimes only a few types of amino acids, for example, two to five types of amino acids (e.g., selected from G, E, D, S, T, A and P), unlike typical proteins and typical linkers which are generally composed of most of the twenty types of amino acids. URPs can contain negatively charged residues (Asp, Glu) at 30, 25, 20, 15, 12, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 percent of the positions. amino acids.

[0112]GRS-containing URPs are of particular interest due, in part, to the high conformational freedom of glycine-containing peptides. Denatured peptides in solution have a high degree of conformational freedom. Most of this conformational freedom is lost after the binding of said peptides to a target such as a receptor, an antibody or a protease. This loss of entropy needs to be compensated by the interaction energy between the peptide and its target. The degree of conformational freedom of a denatured peptide depends on its amino acid sequences. Peptides that contain many amino acids with small side chains tend to have more conformational freedom than peptides that are composed of amino acids with larger side chains. Peptides containing the amino acid glycine have particularly large degrees of freedom. It has been estimated that glycine-containing peptide bonds have approximately 3.4 times more entropy in solution relative to corresponding alanine-containing sequences (D'Aquino, J. A. et al. (1996) Proteins, 25: 143-56). This factor increases with the number of glycine residues in a sequence. As a result, such peptides tend to lose more entropy upon binding to targets, which reduces their overall ability to interact with other proteins, as well as their ability to adopt defined three-dimensional structures. The great conformational flexibility of glycine peptide bonds is also evident when analyzing Ramachandran representations of protein structures in which glycine peptide bonds occupy areas that are rarely occupied by other peptide bonds (Venkatachalam, C. M. et al. (1969). ) Annu Rev Biochem, 38: 45-82). Stites et al. studied a database of 12,320 residues from 61 non-homologous high-resolution crystal structures to determine the preferred phi, psi conformations of each of the 20 amino acids. The distributions observed in the native state of proteins are assumed to also reflect the distributions found in the denatured state. The distributions were used to approximate the surface energy of each residue, allowing the calculation of relative conformational entropies for each residue with respect to glycine. In the most extreme case, the substitution of proline with glycine, conformational entropy changes will stabilize the native state with respect to the denatured state -0.82 /- 0.08 kcal/mol at 20 °C (Stites, W. E. et al. (1995) Protein, 22: 132). These observations confirm the special role of glycine among the 20 natural amino acids.

[0113]Natural or non-natural sequences can be used in the design of the object URPs. For example, a multitude of natural sequences containing high glycine content are provided in Table 1, Table 2, Table 3 and Table 4. A person skilled in the art can adopt any one of the sequences as a URP, or modify the sequences to achieve the intended properties. If immunogenicity is of interest to the host subject, it is preferable to design GRS-containing URPs based on host-derived glycine-rich sequences. Preferred GRS-containing URPs are human protein sequences or sequences that share substantial homology with the corresponding glycine-rich sequences in human reference proteins.

[0114]

Table 1. Structural analysis of proteins containing glycine-rich sequences.

______________

Table 2: Open reading frames encoding GRS with 300 or more glycine residues Access Organism Gly (%) GRS length Gene length Predicted functionNP_974499Arabidopsis thaliana64,509,579 unknown ZP_00458077Burkholderia66,373,518 putativacenocepacia lipoprotein

XP_477841Oryza sativa74 371 422 unknown MP_910409Oryza sativa75 368 400 putative cell wall precursor NP_610660Drosophila66 322 610 transposable elementmelanogaster

Table 3. Examples of human GRS

Gly access (%) GRS length Gene length Hydrophobia Predicted functionNP_000217 62 135 622 yes keratin 9

NP_631961 61 73 592 yes TBP-associated factor 15 isoform 1 NP_476429 65 70 629 yes keratin 3

NP 000418 70 66 316 yes loricrin, cell envelope NP_056932 60 66 638 yes cytokeratin 2

Table 4. Additional examples of human GRS.

Access Sequences Number of amino acidsNP_006228 GPGGGGGPGGGGGPGGGGPGGGGGGGPGGGGGGPGGG 37

NP_005850 RGGGRGGG 22 NP_787059 GAGGGGGGGGGGGGGSGGGGGGGAGAGGAGAG 33 NP_009060 GGGSGSGGAGGGSGGGSGSGGGGGGGAGGGGGG 32 NP_031393 GDGGGAGGGGGGGGSGGGGSGGGGGGG 27 NP_115818 GSGGSGGSGGGPGPGPGGGGG 21 XP_376532 GEGGG GGGEGGGAGGGSG 18

NP 065104 GGGGGGGGDGGG 12 GGGSGSGGAGGGGSGGGSGSGGGGGGAGGGGGGGSSGGGSGTAGGHSG

POU domain, class 4, transcription factor 1 [Homo sapiens] GPGGGGGPGGGGGPGGGGPGGGGGGGPGGGGGGPGGG

2 containing YEATS domain [Homo sapiens]

GGSGAGGGGGGGGGGGSGSGGGGSTGGGGGTAGGG

AT-rich interactive domain 1B form 3 (similar to SWI1); BRG1-binding protein ELD/OSA1; Eld (eyelid)/Osa protein [Homo sapiens]

GAGGGGGGGGGGGGGSGGGGGGGGGAGAGGAGAG

AT-rich interacting domain 1B isoform 2 (similar to SWI1); BRG1-binding protein ELD/OSA1; Eld (eyelid)/Osa protein [Homo sapiens]

GAGGGGGGGGGGGGGSGGGGGGGGGAGAGGAGAG

AT-rich interacting domain 1B isoform 1 (similar to SWI1); BRG1-binding protein ELD/OSA1; Eld (eyelid)/Osa protein [Homo sapiens]

GAGGGGGGGGGGGGGSGGGGGGGGGAGAGGAGAG

purine-rich element binding protein A; purine-rich single-stranded DNA binding protein alpha; transcriptional activator protein PUR-alpha [Homo sapiens]

GHPGSGSGSGGGGGGGGGGGGGSGGGGGGAPGG

regulatory factor X1; trans-acting regulatory factor 1; enhancing factor C; MHC class II regulatory factor RFX [Homo sapiens]

GGGGSGGGGGGGGGGGGGGGSTGGGGSGAG

Bromine domain-containing protein altered in leukemia [Homo sapiens] GGRGRGGRGRGSRGRGGGGTRGRGRGRGGRG

unknown protein [Homo sapiens]

GSGGSGGSGGGGPGPGGGGGPSGSGSGPG PREDICTED: hypothetical protein XP_059256 [Homo sapiens] GGGGGGGGGGGRGGGGRGGGRGGGGEGGG

zinc finger protein 281; ZNP-99 transcription factor [Homo sapiens]

GGGGT GSSGGGSGSGGGGSGGGGGGGSSG

short isoform of RNA-binding protein (autoantigenic, hnRNP-associated with lethal yellow); RNA binding protein (autoantigenic); RNA binding protein (lethal) [Homo sapiens] GDGGGAGGGGGGGGSGGGGGSGGGGGGG

68 kDa signal recognition particle [Homo sapiens] GGGGGGSGGGGGSGGGGSGGGRGAGG

protein KIAA0265 [Homo sapiens]

GGGAAGAGGGGSGAGGGSGSGGRGTG

homologue of engrailed2;Engrailed-2[Homo sapiens]

GAGGGRGGGAGGEGGASGAEGGGGAGG

RNA-binding protein long isoform (autoantigenic, hnRNP-associated with lethal yellow), RNA-binding protein (autoantigenic); RNA-binding protein (autoantigenic, hnRNP-associated with lethal yellow) [Homo sapiens]

GDGGGAGGGGGGGGSGGGGSGGGGGG

androgen receptor; dihydrotestosterone receptor [Homo sapiens] GGGGGGGGGGGGGGGGGGGGGGGEAG

homeotic box D11; homeotic box 4F; Hox-4.6, mouse, homolog of; homeotic box protein Hox-D11 [Homo sapiens]

GGGGGGSAGGGSSGGPGGGGGGGGAGG

frizzled 8; frizzled homolog 8 (Drosophila) [Homo sapiens]

GGGGGPGGGGGGGPGGGGGPGGGGG

gene associated with eye development [Homo sapiens]

GRGGAGGSGGAGGSGAAGGTGSSGGGG

homeotic box B3; 2G homeotic box; homeotic box protein Hox-B3 [Homo sapiens] GGGGGGGGGGGGSGGSGGGGGGGGGG

open reading frame 29 of chromosome 2 [Homo sapiens]

GGSGGGRGGASGPGSGSGGPGGPAG

protein DKFZP564F0522 [Homo sapiens]

PREDICTED GGHHGDRGGGRGGRGGRGGRGGRAG: similar to even-skipped homologous protein 2 (EVX-2) of homeotic box [Homo sapiens] GSRGGGGGGGGGGGGGGGGAGAGGG

ras homologous gene family, member U; Ryu GTPase; Wnt-1-responsive Cdc42 homolog; 2310026M05Rik; GTP-binding protein-like 1; CDC42-like GTPase [Homo sapiens] GGGRGGRGPGEPGGRGRAGGAEGRG

scratch protein 2; transcriptional repressor scratch 2; scratch (Drosophila homolog) 2, zinc finger protein [Homo sapiens]

GGGGGDAGGSGDAGGAGGRAGRAG

nucleolar protein family A, member 1; GAR1 protein [Homo sapiens] GGGRGGRGGGRGGGGRGGGRGGG

keratin 1; keratin-1; cytokeratin 1; alpha-hair protein [Homo sapiens] GGSGGGGGGSSGGRGGSGGGSSGG

hypothetical protein FLJ31413 [Homo sapiens]

GSGPGT GGGGSGSGGGGGGGSGGG

onecut domain, family member 2; onecut 2 [Homo sapiens]

GARGGGSGGGGGGGGGGGGGGPG

POU domain, class 3, transcription factor 2 [Homo sapiens]

GGGGGGGGGGGGGGGGGGGGGDG: similar to complete THO subunit 4 (Tho4) (RNA and export factor binding protein 1) (REF1-I) (Ally of AML-1 and LEF-1) (Aly/REF) [Homo sapiens ]

PREDICTED GGTRGGTRGGTRGGDRGRGRGAG: similar to THO complex subunit 4 (Tho4) (RNA and export factor binding protein 1) (REF1-I) (Ally of AML-1 and LEF-1) (Aly/REF) [Homo sapiens ]

GGTRGGTRGGTRGGDRGRGRGAG

POU domain, class 3, transcription factor 3 [Homo sapiens]

GAGGGGGGGGGGGGGGGAGGGGGG

nucleolar protein family A, member 1; GAR1 protein [Homo sapiens] GGGRGGRGGGRGGGGRGGGRGGG

fibrillarin; nucleolar scleroderma antigen 34 kD; RNA, U3 small nucleolar-interacting protein 1 [Homo sapiens]

GRGRGGGGGGGGGGGGGRGGGG

zinc finger protein 579 [Homo sapiens]

GRGRGRGRGRGRGRGRGRGGAG

calpain, small subunit 1; calcium-activated neutral proteinase; calpain, small polypeptide; calpain 4, small subunit (30 K); calcium-dependent protease, small subunit [Homo sapiens] GAGGGGGGGGGGGGGGGGGGGG

keratin 9 [Homo sapiens]

GGGGSGGGHSGGGSGGGHSGGSGG

D1 forkhead box; forkhead related trigger 4; Forkhead, drosophila, 8 homologous-like; 8 similar to forkhead (Drosophila) [Homo sapiens]

PREDICTED GAGAGGGGGGGGAGGGSAGSG: similar to RIKEN cDNA C230094B15 [Homo sapiens]

GGPGT GSGGGGAGT GGGAGGPG GGGGGGGGGAGGAGGAGSAGGG

cadherin 22 precursor, rat cadherin ortholog PB [Homo sapiens] GGDGGGSAGGGAGGGSGGGAG

AT-binding transcription factor 1; AT motif binding factor 1 [Homo sapiens] GGGGGGSGGGGGGGGGGGGGG

eomesodermin; t-box, brain, 2; homologue of eomesodermina (Xenopus laevis) [Homo sapiens] GPGAGAGSGAGGSSGGGGGPG

membrane-associated phosphatidylinositol transfer protein, 2; PYK2 N-terminal domain-interacting receptor 3; B alpha 2 of retinal degeneration (Drosophila) [Homo sapiens] GGGGGGGGGGGSSGGGGSSGG

sperm-associated antigen 8 isoform 2; sperm membrane protein 1 [Homo sapiens]

PREDICTED GSGSGPGPGSGPGSGPGHGSG: RNA-binding motif protein 27 [Homo sapiens]

GPPGPGPGPGPGPGPGPGPG

AP1 gamma subunit binding protein 1 isoform 1; gamma-synergine complex 1 gamma subunit binding protein 1; adapter-related protein [Homo sapiens] GAGSGGGGAAGAGAGSAGGGG

AP1 gamma subunit binding protein 1 isoform 2; gamma-synergine; Adapter-related protein complex 1 gamma subunit binding protein 1 [Homo sapiens] GAGSGGGGAAGAGAGSAGGGG

ankyrin repeat and sterile alpha motif domain-containing 1; ankyrin repeat and SAM domain-containing 1 [Homo sapiens]

GGGGGGGGSGGGGGGGSGGGGGG

Methyl-CpG binding domain protein 2 isoform 1 [Homo sapiens] GRGRRGRGRGRGRGRGRGRGRG

triple functional domain (interacting with PTPRF) [Homo sapiens]

GGGGGGGGSGGGSGGGGGSGGGG

forkhead box D3 [Homo sapiens]

GGEEGGASGGGPGAGSGSAGG

sperm-associated antigen 8 isoform 1; sperm membrane protein 1 [Homo sapiens]

GSGSGPGPGSGPGSGPGHGSG

testis-specific isoform of methyl-CpG binding domain protein 2 [Homo sapiens] GRGRRGRGRGRGRGRGRGRGRG

aven cell death regulator; programmed cell death 12 [Homo sapiens] GGGGGGGGDGGGRRGRRGRGRG

termination transcript regulator 1; delta helicase; homolog of Up-Frameshift mutation 1 (S. cerevisiae); termination mRNA reducing factor 1; yeast homolog Upflp [Homo sapiens] GGPGGPGGGGAGGPGGAGAG

small-conductance calcium-activated potassium channel protein 2 isoform a; apamin-sensitive small-conductance Ca2+-activated potassium channel [Homo sapiens]

GT GGGGST GGGGGGGSGHG

SRY box 1 (sex determining region Y); SRY-related HMG box gene 1 [Homo sapiens] GPAGAGGGGGGGGGGGGGGG

transcription factor 20 isoform 2; stromelysin platelet-derived growth factor-responsive element binding protein-1; stromelysin PDGF-responsive element binding protein 1; SPRE binding protein; nuclear SPBP factor [Homo sapiens]

GGTGGSSGSSSGSGSGGGRRG

transcription factor 20 isoform 1; stromelysin platelet-derived growth factor-responsive element binding protein-1; stromelysin PDGF-responsive element binding protein 1; SPRE binding protein; nuclear SPBP factor [Homo sapiens]

GGTGGSSGSSSGSGSGGGRRG

ras-interacting protein 1 [Homo sapiens]

GSGT GTT GSSGAGGPGTPGG

BMP-2-inducible kinase isoform b [Homo sapiens]

GGSGGGAAGGGAGGAGAGAG

BMP-2-inducible kinase isoform a [Homo sapiens]

GGSGGGAAGGGAGGAGAGAG

C1 forkhead box; forkhead related trigger 3; Forkhead, drosophila, 7 homologous-like; 7 forkhead-like (Drosophila); iridogoniodysgenesis type 1 [Homo sapiens]

GSSGGGGGGGAGAAGGAGGAG

p54 splicing factor; 54 kDa arginine-rich nuclear protein [Homo sapiens] GPGPSGGPGGGGGGGGGGGGG

v-maf musculoaponeurotic fibrosarcoma oncogenic homologue; avian musculoaponeurotic fibrosarcoma (MAF) proto-oncogene; oncogenic homologue of v-maf musculoaponeurotic fibrosarcoma (avian) [Homo sapiens]

GGGGGGGGGGGGGGGAAGAGG

small nuclear ribonucleoprotein D1 polypeptide 16 kDa; snRNP core protein D1; Sm-D autoantigen; small nuclear ribonucleoprotein D1 polypeptide (16kD) [Homo sapiens] GRGRGRGRGRGRGRGRGRGG

hypothetical protein H41 [Homo sapiens]

GSAGGSSGAAGAAGGGAGAG

URP-containing non-glycine residues (NGR):

[0115]The sequences of non-glycine residues in these GRS can be selected to optimize URP properties and hence proteins containing the desired URPs. For example, URP sequences can be optimized to enhance the selectivity of the resulting protein for a particular tissue, specific cell type or cell lineage. For example, protein sequences may be incorporated that are not ubiquitously expressed, but are differentially expressed in one or more of the body's tissues including heart, liver, prostate, lung, kidney, bone marrow, blood, skin, bladder, brain. , muscles, nerves and selected tissues that are affected by diseases such as infectious diseases, autoimmune disease, kidney, neuronal, cardiac disorders and cancers. Sequences representative of a specific developmental origin, such as those expressed in an embryo or an adult, may be used during the formation of ectoderm, endoderm or mesoderm in a multicellular organism. Sequences that participate in a specific biological process including, but not limited to, cell cycle regulation, cell differentiation, apoptosis, chemotaxis, cell motility, and cytoskeletal rearrangement may also be used. Other non-ubiquitously expressed protein sequences may also be used to direct the resulting protein to specific subcellular locations: extracellular matrix, nucleus, cytoplasm, cytoskeleton, plasma and/or intracellular membranous structures including, but not limited to, coated pits, Golgi, endoplasmic reticulum, endosome, lysosome and mitochondria.

[0116]A variety of these tissue-specific, cell type-specific, subcellular localization-specific sequences are known and are available from numerous protein databases. Such selective URP sequences can be obtained by generating libraries of random or semi-random URP sequences, injecting them into animals or patients, and determining sequences with the desired tissue selectivity in tissue samples. Sequence determination can be performed by mass spectrometry. Using similar procedures, URP sequences can be selected that facilitate oral, buccal, intestinal, nasal, thecal, peritoneal, pulmonary, rectal, or dermal uptake.

[0117] Of particular interest are URP sequences that contain regions that are relatively rich in the positively charged amino acids arginine or lysine that promote cellular uptake or transport across membranes. URP sequences can be designed that contain one or more protease-sensitive sequences. Such URP sequences can be cleaved once the product of the method of the invention has reached its target location. This cleavage may trigger an increase in the potency of the pharmaceutically active domain (pro-drug activation) or may enhance binding of the cleavage product to a receptor. URP sequences can be designed to carry excess negative charges by introducing aspartic acid or glutamic acid residues. Of particular interest are URPs that contain more than 5%, more than 6%, 7%, 8%, 9%, 10%, 15%, 30% or more glutamic acid and less than 2% lysine or arginine. Such URPs carry an excess of negative charge and as a result have a tendency to adopt open conformations due to electrostatic repulsion between individual negative charges of the peptide. Such an excess of negative charge leads to an effective increase in their hydrodynamic radius and as a result may lead to reduced renal clearance of such molecules. Thus, the effective net charge and hydrodynamic radius of a URP sequence can be modulated by controlling the frequency and distribution of negatively charged amino acids in the URP sequences. Most tissues and surfaces in a human or animal carry excess negative charges. By designing URP sequences to carry excess negative charges, non-specific interactions between the resulting protein comprising the URP and various surfaces such as blood vessels, healthy tissues or various receptors can be minimized.

[0118]URPs may have a repetitive amino acid sequence of the format (motif) in which one sequence motif forms a direct repeat (i.e. ABCABCABCABC) or an inverted repeat (ABCCBAABCCBA) and the number of these repeats may be 2 , 3, 4, 5, 6, 7, 8, 9, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 35, 40, 50 or more URP or repetitions within URPs frequently contain only 1, 2, 3, 4, 5, or 6 different types of amino acids. URPs typically consist of repeats of human amino acid sequences that have 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 22, 24 , 26, 28, 30, 32, 34, 36 or more amino acids in length, but URPs can also consist of non-human amino acid sequences that are 20, 22, 24, 26, 28, 30, 32, 34, 36, 3840, 42, 44, 46, 48, 50 amino acids long.

URPs derived from human sequences:

[0119]URPs can be derived from human sequences. The human genome contains many subsequences that are rich in a particular amino acid. Of particular interest are those amino acid sequences that are rich in a hydrophilic amino acid such as serine, threonine, glutamate, aspartate or glycine. Of particular interest are those subsequences that contain few hydrophobic amino acids. Such subsequences are predicted to be unstructured and highly soluble in aqueous solution. Such human subsequences can be modified to further improve their usefulness. Figure 17 shows an exemplary human sequence that is rich in serine and that can be isolated as a target URP. The exemplified dentin sialophosphoprotein contains a subsequence of 670 amino acids in which 64% of the residues are serine and most of the other positions are hydrophilic amino acids such as aspartate, asparagines and glutamate. The sequence is extremely repetitive and as a result has a low information content. Subsequences of such a human protein can be used directly. If desired, the sequence can be modified in a way that preserves its overall character, but makes it more suitable for pharmaceutical applications. Examples of sequences that are related to dentin sialophosphoprotein are (SSD)n, (SSDSSN)n, (SSE)n in which n is between approximately 4 and 200.

[0120]The use of human protein sequences is particularly desirable in the design of URPs with reduced immunogenicity in a human subject. A key step in eliciting an immune response to a foreign protein is the presentation of peptide fragments of that protein by MHC class II receptors. These MHCII-bound fragments can then be detected by T cell receptors, which trigger the proliferation of T helper cells and initiate an immune response. Removal of T cell epitopes from pharmaceutical proteins has been recognized as a means of reducing the risk of provoking an immune reaction (Stickler, M. et al. (2003) J Immunol Methods, 281: 95-108). MHCII receptors normally interact with an epitope having, for example, a 9 amino acid long region of the displayed peptides. Thus, the risk of eliciting an immune response to a protein in patients may be reduced if all or most of the possible 9-mer subsequences of the protein can be found in human proteins and, if so, these sequences and repeats of these sequences will not be recognized by the patient as strange sequences. Human sequences can be incorporated into the design of<u>R<p>sequences by oligomerizing or concatenating human sequences that have suitable amino acid compositions. These can be direct repetitions or inverted repetitions or mixtures of different repetitions. For example, sequences can be oligomerized as shown in Table 2. Such oligomers have a reduced risk of being immunogenic. However, the splicing sequences between the monomer units can still contain T cell epitopes that can trigger an immune reaction, which is illustrated in Figure 3. Additionally, the risk of provoking an immune response can be reduced by designing URP sequences based on multiple overlapping human sequences. This approach is illustrated in Figure 4. The URP sequence in Figure 2 was designed as an oligomer based on multiple human sequences so that each 9-mer subsequence of the oligomer can be found in a human protein. In these designs, each 9-mer subsequence is a human sequence. An example of a URP sequence based on three human sequences is shown in Figure 5. It is also possible to design URP sequences based on a single human sequence such that all possible 9-mer subsequences in the oligomeric URP sequences occur. in the same human protein. An example is shown in Figure 6 based on the POU domain that is rich in glycine and proline. The repeat monomer in the URP sequence is only a fragment of the human protein and its flanking sequences are identical to the repeat unit as illustrated in Figure 6. Non-oligomeric URP sequences can also be designed based on human proteins. The primary conditions are that all 9-mer subsequences can be found in human sequences. The amino acid composition of the sequences preferably contain few hydrophobic residues. Of particular interest are URP sequences that are designed based on human sequences and that contain a large fraction of glycine residues.

[0121]Using this scheme or a similar one, a class of URPs can be designed that comprise repeat sequences with low immunogenicity for the host of interest. The host of interest can be any animal, including vertebrates and invertebrates. Preferred hosts are mammals such as primates (e.g. chimpanzees and humans), cetaceans (e.g. whales and dolphins), bats (e.g. bats), perissodactyls (e.g. horses and rhinos), rodents (e.g. rats). ), and certain types of insectivores such as shrews, moles and hedgehogs. If the human is selected as the host, URPs typically contain multiple copies of the sequences or repeat units, in which the majority of the segments comprising about 6 to about 15 contiguous amino acids are present in one or more native human proteins. URPs can also be designed in which the majority of the segments comprising between about 9 and about 15 contiguous amino acids are found in one or more native human proteins. As used herein, most segments refer to more than about 50%, preferably 60%, preferably 70%, preferably 80%, preferably 90%, preferably 100%. If desired, each of the possible segments between about 6 and 15 amino acids, preferably between about 9 and 15 amino acids within the repeat units, are present in one or more native human proteins. URPs may comprise multiple repeat units or sequences, for example, having 2, 3, 4, 5, 6, 7, 8, 9, 10 or more repeat units.

Design URPs that are substantially free of human T cell epitopes:

[0122]URP sequences can be designed to be substantially free of epitopes recognized by human T lymphocytes. For example, one can synthesize a series of semi-random sequences with amino acid compositions that favor denatured unstructured conformations and evaluate these sequences for the presence of human T cell epitopes and whether they are human sequences or not. Assays for human T cell epitopes have been described (Stickler, M. et al. (2003) J Immunol Methods, 281: 95-108). Of particular interest are peptide sequences that can oligomerize without generating T cell epitopes or non-human sequences. This can be achieved by testing direct repeats of these sequences for the presence of T cell epitopes and for the appearance of non-human 6- to 15-mer and particularly 9-mer subsequences. An alternative is to evaluate multiple peptide sequences that can be assembled into repeat units as described in the previous section for assembly of human sequences. Another alternative is to design URP sequences that produce low scores using epitope prediction algorithms such as TEPITOPE (Sturniolo, T. et al. (1999) Nat Biotechnol, 17: 555-61). Another approach to avoiding T cell epitopes is to avoid amino acids that can serve as anchor residues during MHC display of peptides, such as M, I, L, V, F. Hydrophobic amino acids and positively charged amino acids can frequently serve as such residues. anchoring and minimizing its frequency in a URP sequence reduces the probability of generating T lymphocyte epitopes and thus provoking an immune reaction. The selected URPs generally contain subsequences that are found in at least one human protein, and have a lower content of hydrophobic amino acids.

[0123]URP sequences can be designed to optimize protein production. This can be achieved by avoiding or minimizing the repetitiveness of the coding DNA. URP sequences such as poly-glycine may have very desirable pharmaceutical properties, but their manufacture may be difficult due to the high GC content of GRS-encoding DNA sequences and due to the presence of repeating DNA sequences that can lead to recombination. .

[0124]As noted above, URP sequences can be designed to be highly repetitive at the amino acid level. As a result, URP sequences have very low information content and may reduce the risk of provoking an immune reaction.

[0125] Non-limiting examples of URPs containing repeating amino acids are: poly-glycine, poly-glutamic acid, poly-aspartic acid, poly-serine, poly-threonine, (GX)n in which G is glycine and serine, aspartic acid, glutamic acid, threonine or proline and n is at least 20, (GGX)n in which X is serine, aspartic acid, glutamic acid, threonine or proline and n is at least 10, (GGGGX)n in which where X is serine, aspartic acid, glutamic acid, threonine or proline, n is at least 15, and z is between 1 and 20.

[0126]The number of these repeats can be any number between 10 and 100. The products of the methods of the invention can contain URP sequences that are semi-random sequences. Examples are semi-random sequences containing at least 30, 40, 50, 60 or 70% glycine in which the glycines are well dispersed and in which the total concentration of tryptophan, phenylalanine, tyrosine, valine, leucine and Isoleucine is then less than 70, 60, 50, 40, 30, 20 or 10% when combined. A preferred semi-random URP sequence contains at least 40% glycine and the total concentration of tryptophan, phenylalanine, tyrosine, valine, leucine and isoleucine is then less than 10%. A more preferred random URP sequence contains at least 50% glycine and the total concentration of tryptophan, phenylalanine, tyrosine, valine, leucine and isoleucine is then less than 5%. URP sequences can be designed by combining the sequences of two or more shorter URP sequences or fragments of URP sequences. Such a combination allows better modulation of the pharmaceutical properties of the product containing the URP sequences and allows the repetitiveness of the DNA sequences encoding the URP sequences to be reduced, which can improve the expression and reduce recombination of URP coding sequences. .

[0127]URP sequences can be designed and selected to possess several of the following desired properties: a) high genetic stability of the coding sequences in the production host, b) high level of expression, c) low immunogenicity (predicted/calculated ), d) high stability in the presence of serum proteases and/or other tissue proteases, e) large hydrodynamic radius under physiological conditions. An exemplary approach to obtaining URP sequences that meet multiple criteria is to construct a library of candidate sequences and identify suitable subsequences from the library. The libraries may comprise random and/or semi-random sequences. Of particular use are codon libraries, which are a library of DNA molecules that contain multiple codons for the identical amino acid residue. Codon randomization can be applied to selected amino acid positions of a certain type or to most or all positions. True codon libraries only encode a single amino acid sequence, but can be easily combined with amino acid libraries, which is a population of DNA molecules that encode a mixture of amino acids (related or unrelated) at the residue position. Codon libraries allow the identification of genes that have relatively low repeatability at the DNA level, but that encode highly repetitive amino acid sequences. This is useful because repetitive DNA sequences tend to recombine, leading to instability. Codon libraries encoding limited amino acid diversity can also be constructed. Such libraries allow the introduction of a limited number of amino acids at some positions in the sequence, while other positions allow codon variation, but all codons encode the same amino acid. Partially random oligonucleotides can be synthesized by incorporating mixtures of nucleotides at the same position during oligonucleotide synthesis. Such partially random oligonucleotides can be fused by overlap PCR or ligation-based approaches. In particular, semi-random oligonucleotides encoding glycine-rich sequences can be multimerized. These oligonucleotides can differ in length and sequences and codon usage. As a result, a library of candidate URP sequences is obtained. Another procedure to generate libraries is to synthesize a starting sequence and subsequently subject these sequences to partial randomization. This can be done by culturing the gene encoding the URP sequences in a mutator strain or by amplification of the encoding gene under mutagenic conditions (Leung, D. et al. (1989) Technique, 1: 11-15). URP sequences with desirable properties can be identified from libraries using a variety of procedures. Sequences that have a high degree of genetic stability can be enriched by growing the library in a production host. Sequences that are unstable will accumulate mutations, which can be identified by DNA sequencing. URP sequence variants that can be expressed at high level can be identified by screening or selection using multiple protocols known to anyone skilled in the art. For example, multiple strains isolated from a library can be grown and expression levels compared. Expression levels can be measured by gel analysis, analytical chromatography or various ELISA-based procedures. Determination of expression levels of individual sequence variants can be facilitated by merging the library of candidate URP sequences with sequence tags such as myc tag, His tag, HA tag. Another approach is to fuse the library with an enzyme or other reporter protein such as green fluorescent protein. Of particular interest is the fusion of the library with a selection marker such as beta-lactamase or kanamycin-acyl transferase. Antibiotic selection can be used to enrich for variants with high expression level and good genetic stability. Variants with good protease resistance can be identified by screening for intact sequences after incubation with proteases. An effective way to identify protease-resistant URP sequences is bacterial phage display or related display procedures. Multiple systems have been described in which sequences undergoing rapid proteolysis can be enriched by phage display. These procedures can be easily adopted to enrich for protease-resistant sequences. For example, a library of candidate URP sequences can be cloned between an affinity tag and the pIII protein of phage M13. The library can then be exposed to proteases or protease-containing biological samples such as blood or lysosomal preparations. Phage-containing protease-resistant sequences can be captured after protease treatment by binding to the affinity tag. Sequences that resist degradation by lysosomal preparations are of particular interest because lysosomal degradation is a key step during antigen presentation in dendritic cells and other antigen-presenting cells. Phage display can be used to identify candidate URP sequences that do not bind a particular immune serum in order to identify URP sequences with low immunogenicity. Animals can be immunized with a candidate URP sequence or with a library of URP sequences to produce antibodies against the URP sequences in the library. The resulting serum can then be used for phage immunopurification to remove or identify sequences that are recognized by antibodies in the resulting immune serum. Other procedures such as bacterial display, yeast display, ribosomal display can be used to identify URP sequence variants with desirable properties. Another approach is the identification of URP sequences of interest by mass spectrometry. For example, a library of candidate URP sequences can be incubated with a protease or biological sample of interest and sequences that resist degradation by mass spectrometry can be identified. In a similar approach, URP sequences that facilitate oral uptake can be identified. A mixture of candidate URP sequences can be fed to animals or humans and variants with the highest efficiency of transfer or uptake across some barrier tissue (i.e., dermal, etc.) can be identified by mass spectrometry. In a similar way, URP sequences that favor other uptake mechanisms such as pulmonary, intranasal, rectal, and transdermal administration can be identified. URP sequences that favor cellular uptake or URP sequences that resist cellular uptake can also be identified.

[0128]URP sequences can be designed by combining URP sequences or fragments of URP sequences that were designed by any of the procedures described above. Furthermore, semi-random approaches can be applied to optimize sequences that were designed based on the rules described above. Of particular interest is codon optimization with the aim of improving the expression of enhanced proteins and improving the genetic stability of the coding gene in production hosts. Codon optimization is of particular importance for URP sequences that are rich in glycine or have highly repetitive amino acid sequences. Codon optimization can be performed using computer programs (Gustafsson, C. et al. (2004) Trends Biotechnol, 22: 346-53), some of which minimize ribosomal arrest (Coda Genomics Inc.). Several properties can be considered when designing URP sequences. Repetitivity in the coding DNA sequences can be minimized. Additionally, the use of codons that are rarely used by the production host (i.e., the AGG and AGA arginine codons and a leucine codon in E. coli) can be avoided or minimized. Glycine tend to have a high GC content which can lead to instability or low expression levels. Thus, where possible, it is preferred to choose codons so that the GC content of the URP coding sequence is suitable for the production organism that will be used to manufacture the URP.

[0129]Genes encoding URPs can be prepared in one or more steps, either completely synthetically or by synthesis combined with enzymatic procedures such as restriction enzyme-mediated cloning, PCR and overlap extension. URP modules can be constructed such that the gene encoding the URP module has low repeatability, while the encoded amino acid sequence has a high degree of repeatability. The approach is illustrated in Figure 11. As a first stage, a library of relatively short URP sequences is constructed. This may be a pure codon library so that each member of the library has the same amino acid sequence, but many different coding sequences are possible. To facilitate the identification of well-expressing library members, the library can be constructed as a fusion with a reporter protein. Examples of suitable reporter genes are green fluorescent protein, luciferase, alkaline phosphatase, beta-galactosidase. Short URP sequences that can be expressed in high concentration in the host organism of choice can be identified by screening. Subsequently, a library of random URP dimers can be generated and the screening for high level expression can be repeated. Dimerization can be performed by ligation, overlap extension or similar cloning techniques. This dimerization and subsequent screening procedure can be repeated multiple times until the resulting uRp sequence has reached the desired length. Optionally, clones in the library can be sequenced to eliminate isolates containing undesirable sequences. The initial library of short URP sequences may allow for some variation in the amino acid sequence. For example, some codons can be randomized so that several hydrophilic amino acids can be produced at that position. During the iterative multimerization procedure, library members can be screened for other characteristics such as solubility or protease resistance, in addition to screening for high-level expression. Instead of dimerizing URP sequences, longer multimers can also be generated. This allows the length of URP modules to be increased more quickly.

[0130]Many URP sequences contain particular amino acids at a high fraction. Such sequences may be difficult to produce by recombinant techniques, since their coding genes may contain repetitive sequences that are subject to recombination. Furthermore, genes containing particular codons at very high frequencies can limit expression, as the respective charged tRNAs in the production host become limiting. An example is the recombinant production of GRS. Glycine residues are encoded by 4 triplets, GGG, GGC, GGA and GGT. As a result, genes encoding GRS tend to be high in Gc and tend to be particularly repetitive. An additional challenge may result from production host codon bias. In the case of E. coli, two glycine codons, GGA and GGG, are rarely used in highly expressed proteins. Thus, codon optimization of the gene encoding URP sequences may be highly desirable. Codon usage can be optimized using software that considers production host codon bias (Gustafsson, C. et al. (2004) Trends Biotechnol, 22: 346-53). Alternatively, codon libraries can be constructed in which all members of the library encode the same amino acid sequence, but in which the codon usage is varied. Such libraries can be screened for highly expressing and genetically stable members that are particularly suitable for large scale production of URP-containing products.

Multivalent Unstructured Recombinant Proteins (MURP):

[0131]As noted above, the subject URPs are particularly useful as modules for designing proteins of therapeutic value. Accordingly, proteins comprising one or more subject URPs are disclosed herein. Such proteins are referred to herein as multivalent unstructured recombinant proteins (MURP).

[0132]To construct MURP, one or more URP sequences can be fused to the N-terminus or C-terminus of a protein or inserted into the center of the protein, for example, in loops of a protein or between modules of the protein of interest. , giving the resulting improved properties of modified proteins with respect to unmodified proteins. The combined length of URP sequences that are bound to a protein can be 200 or more amino acids.

[0133]MURPs have one or more improved properties as detailed below.

Improved half-life:

[0134]Adding URP sequences to a pharmaceutically active protein can improve many properties of that protein. In particular, adding a long URP sequence can significantly increase the serum half-life of the protein. Such URPs contain amino acid sequences of at least about 200 or more amino acids.

[0135]URPs can be fragmented so that the resulting protein contains multiple URPs, or multiple fragments of URPs. In one aspect, fused URPs can increase the hydrodynamic radius of a protein and thus reduce its clearance from the blood by the kidney. The increase in the hydrodynamic radius of the resulting fusion protein relative to the unmodified protein can be detected by ultracentrifugation, size exclusion chromatography, or light scattering.

Improved tissue selectivity:

[0136]Increased hydrodynamic radius can also lead to reduced tissue penetration, which can be exploited to minimize side effects of a pharmaceutically active protein. It is well documented that hydrophilic polymers have a tendency to accumulate in tumor tissue that is produced by the enhanced permeability and retention (EPR) effect. The underlying cause of the EPR effect is the porous nature of the tumor vasculature (McDonald, D. M. et al. (2002) Cancer Res, 62: 5381-5) and the lack of lymphatic drainage in tumor tissues. Therefore, the selectivity of pharmaceutically active proteins for tumor tissues can be enhanced by adding hydrophilic polymers. As such, the therapeutic index of a given pharmaceutically active protein can be increased by incorporating the subject URPs.

Protection from degradation and reduced immunogenicity:

[0137]The addition of URP sequences can significantly improve the protease resistance of a protein. The URP sequences themselves can be designed to be resistant to proteases and binding them to a protein can protect that protein from access by degrading enzymes. URP sequences can be added to pharmaceutically active proteins with the aim of reducing undesirable interactions of the protein with other receptors or surfaces. To achieve this it may be beneficial to add the URP sequences to the pharmaceutically active protein in proximity to the site on the protein that makes such unwanted contacts. In particular, URP sequences can be added to pharmaceutically active proteins with the aim of reducing their interactions with any component of the immune system to prevent an immune response against the product of the method of the invention. The addition of a URP sequence to a pharmaceutically active protein can reduce the interaction with preexisting antibodies or B cell receptors. In addition, the addition of URP sequences can reduce the uptake and processing of the product of the method of the invention by presenting cells. antigen. The addition of one or more URP sequences to a protein is a preferred way to reduce its immunogenicity, as it will suppress an immune response in many species, allowing the expected immunogenicity of a product in patients to be predicted based on animal data. Such species-independent testing of immunogenicity is not possible for approaches that rely on the identification and deletion of human T cell epitopes or comparison of sequences with human sequences.

Disruption of T cell epitopes:

[0138]URP sequences can be introduced into proteins in order to disrupt T cell epitopes. This is particularly useful for proteins that combine multiple separate functional modules. The formation of T cell epitopes requires that peptide fragments of a protein antigen bind to MHC. MHC molecules interact with a short amino acid segment, typically 9 contiguous residues of the presented peptides. Direct fusion of different binding modules on one protein molecule can lead to T cell epitopes that span two neighboring domains. Separating functional modules by URP modules prevents the generation of such module-spanning T cell epitopes as illustrated in Figure 7. Insertion of URP sequences between functional modules may also interfere with proteolytic processing in antigen-presenting cells. , which will lead to a further reduction in immunogenicity. Another approach to reduce the risk of immunogenicity is to alter T cell epitopes within functional modules of a product. In the case of microproteins, one approach is for some of the intercysteine loops (those not involved in target binding) to be rich in glycine. In microproteins, whose structure is due to a small number of cysteines, most or all of the residues that do not participate in target binding could in fact be replaced with glycine, serine, glutamate, threonine, thus reducing the possibilities of immunogenicity, to the while the affinity for the target is not affected. For example, this can be carried out by performing a 'glycine scan' of all residues, in which each residue is replaced with a glycine, then selecting clones that retain target binding using display or phage screening, and then combining all glycine substitutions that are allowed. In general, functional modules are much more likely to contain T cell epitopes than URP modules. The frequency of T cell epitopes in functional modules can be reduced by replacing all or many non-critical amino acid residues with small hydrophilic residues such as gly, ser, ala, glu, asp, asn, gln, thr. Positions in a functional module that allow substitution can be identified using a variety of random or structure-based protein engineering approaches.

Improved solubility:

[0139]The functional modules of a protein may have limited solubility. In particular, binding modules tend to carry hydrophobic residues on their surface, which can limit their solubility and can lead to aggregation. Separating or flanking such functional modules with URP modules can improve the overall solubility of the resulting product. This is particularly true for URP modules that carry a significant percentage of hydrophilic or charged residues. Separating functional modules with soluble URP modules can reduce intramolecular interactions between these functional modules.

Improved pH profile and product loading homogeneity:

[0140]URP sequences can be designed to carry excess negative or positive charges. As a result, they confer an electrostatic field to any fusion component that can be used to shift the pH profile of an enzyme or binding interaction. Furthermore, the electrostatic field of a charged URP sequence can increase the homogeneity of pKa values of surface charges of a protein product, leading to abrupt pH profiles of ligand interactions and to refined separations by isoelectrofocusing or chromatofocusing.

Improved purification properties due to more refined product pKa:

[0141]Each amino acid in solution itself has a unique fixed pKa, which is the pH at which its functional groups are half protonated. In a typical protein you have many types of residues and due to the proximity and respiration effects of the protein, they also change each other's effective pKa in variable ways. Because of this, at a wide range of pH conditions, typical proteins can adopt hundreds of differently ionized species, each with a different molecular weight and net charge, due to large numbers of combinations of charged and neutral amino acid residues. This is called a broad ionization spectrum and makes the analysis (i.e., mass spectrometry) and purification of such proteins more difficult.

[0142]PEG has no charge and does not affect the ionization spectrum of the protein to which it is bound, leaving it with a broad ionization spectrum. However, a URP with a high content of Gly and Glu only exists in principle in two states: neutral (-COOH) if the pH is below the pKa of glutamate and negatively charged (-COO') if the pH is above of the pKa of glutamate. URP modules can form a single type of homogeneously ionized molecule and can give a single mass in mass spectrometry.

[0143]If desired, MURPs can be expressed as a fusion with a URP that has a single type of charge (Glu) distributed at constant spacing across the URP module. One can choose to incorporate 25-50 residues of Glu per 20 kD of URP and all of these 25-50 residues would have very similar pKa.

[0144]In addition, the addition of 25-50 negative charges to a small protein such as IFN, hGH or GCSF (with only 20 charged residues) would increase the charge homogeneity of the product and increase its isoelectric point, which would be very close to the pKa of free glutamate.

[0145]The increase in charge homogeneity of the protein population has favorable processing properties such as in ion exchange, isoelectric focusing, mass spectrometry, etc., compared to traditional PEGylation.

Improved formulation and/or administration:

[0146]The addition of URP sequences to pharmaceutically active proteins can significantly simplify the formulation and/or administration of the resulting products. URP sequences can be designed to be very hydrophilic and as a result improve the solubility of (for example) human proteins, which often contain hydrophobic patches that are used to bind to other human proteins. The formulation of such human proteins, such as antibodies, can be quite demanding and often limits their concentration and delivery options. URPs can reduce product precipitation and aggregation and allow simpler formulations to be used that contain fewer components, which are normally necessary to stabilize a product in solution. The improved solubility of products containing URP sequences allows these products to be formulated at higher concentrations and as a result the injection volume for injectable products can be reduced, which may allow injection at home, which is limited to a very low injected volume. The addition of a URP sequence can also simplify storage of the resulting formulated products. URP sequences can be added to pharmaceutically active proteins to facilitate their oral, pulmonary, rectal or intranasal uptake. URP sequences can facilitate various modes of administration because they allow for higher product concentrations and improved product stability. Additional improvements can be achieved by designing URP sequences that facilitate membrane penetration. Improved production:

[0147]The addition of URP sequences can have significant benefits to the production of the resulting product. Many recombinant products, especially native human proteins, have a tendency to form aggregates during production that can be difficult or impossible to dissolve and even when removed from the final product can occur again. These are normally due to hydrophobic patches by which these proteins (native human) came into contact with other proteins (native human) and mutation of these residues is considered risky due to immunogenicity. However, URPs can increase the hydrophilicity of such proteins and allow their formulation without mutating the human protein sequence. URP sequences can facilitate the folding of a protein to reach its native state. Many pharmaceutically active proteins are produced by recombinant procedures in a non-native aggregated state. These products need to be denatured and subsequently incubated under conditions that allow the proteins to fold into their native active state. A common side reaction during renaturation is the formation of aggregates. Fusion of URP sequences with a protein significantly reduces its tendency to form aggregates and thus facilitates the folding of the pharmaceutically active component of the product. Products containing URP are much easier to prepare relative to polymer-modified proteins. Modification with chemical polymers requires additional modification and purification steps after the active protein is purified. In contrast, URP sequences can be manufactured using recombinant DNA procedures together with the pharmaceutically active protein. The products of the processes of the invention are also significantly easier to characterize compared to polymer modified products. Due to the recombinant production procedure, more homogeneous products with defined molecular characteristics can be obtained. URP sequences can also facilitate purification of a product. For example, URP sequences may include subsequences that can be captured by affinity chromatography. An example is sequences rich in histidine, which can be captured on resins with immobilized metals such as nickel. URP sequences can also be designed to have an excess of negatively or positively charged amino acids. As a result, they can significantly impact the net charge of a product, which can facilitate purification of products by ion exchange chromatography or preparative electrophoresis.

[0148]Object MURPs may contain a variety of modules including, but not limited to, binding modules, effector modules, multimerization modules, C-terminal modules, and N-terminal modules. Figure 1 represents a MURP as example that has multiple modules. However, MURPs can also have relatively simple architectures which are illustrated in Fig. 2. MURPs can also contain fragmentation sites. These may be protease-sensitive sequences or chemically sensitive sequences that can be preferentially cleaved when MURPs reach their target site.

Bonding Module (BM):

[0149]The MURPs used in the methods of the present invention may comprise one or more binding modules. Binding module (BM) refers to a peptide or protein sequence that can specifically bind to one or more targets, which may be one or more therapeutic targets or accessory targets, such as for targeting cells, tissues or organs. BMs may be linear or cyclic peptides, cysteine-restricted peptides, microproteins, scaffolding proteins (e.g., fibronectin, ankyrins, crystallin, streptavidin, antibody fragments, domain antibodies), peptide hormones, growth factors, cytokines, or any type of protein domain, human or non-human, natural or non-natural, and may be based on a natural scaffold or not based on a natural scaffold, or based on combinations or may be fragments of any of the above. Optionally, these BMs can be manipulated by adding, removing or substituting one or multiple amino acids in order to enhance their binding properties, their stability, or other properties. Binding modules can be obtained from natural proteins, by design, or by gene package display, including phage display, cell display, ribosomal display, or other display procedures. The binding modules can bind to the same copy of the same target, which produces avidity, or they can bind to different copies of the same target (which can produce avidity if these copies are connected or linked in some way, such as by a membrane cellular), or they can bind to two unrelated targets (which gives avidity if these targets are linked in some way, such as by a membrane). Binding modules can be identified by screening or otherwise analyzing random libraries of peptides or proteins.

[0150]Particularly desirable binding modules are those that upon incorporation into a MURP, the MURP gives a desirable T epitope score. The T epitope score of a protein is the log Kd (dissociation constant, affinity, dissociation rate) of the binding of that protein to multiple of the most common human MHC alleles, as disclosed in Sturniolo T. et al. . (1999) Nature Biotechnology 17:555). The score ranges over at least 15 logarithms, approximately 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0, -1, -2, -3, -4, -5 (10e10 Kd ) to approximately -5. Preferred MURPs give a lower score of approximately -3.5 [KKW: On an absolute scale?]

[0151]Also of particular interest are binding modules comprising disulfide bonds formed by pairing of two cysteine residues. In certain embodiments, the binding modules comprise polypeptides that have high cysteine content or high disulfide density (HDD). HDD family bonding modules typically have 5-50% (5, 6, 7, 8, 9, 10, 12, 14, 16, 18, 20, 25, 30, 35, 40, 45 or 50%) of cysteine residues and each domain typically contains at least two disulfides and optionally a co-factor such as calcium or another ion.

[0152]The presence of HDD scaffolding allows these modules to be small, but still adopt a relatively rigid structure. Stiffness is important to obtain high binding affinities, resistance to proteases and heat, including proteases involved in antigen processing, and thus contributes to the low immunogenicity or non-immunogenicity of these modules. The disulfide structural region folds the modules without the need for a large number of hydrophobic side chain interactions in the interior of most modules. The small size is also advantageous for rapid tissue penetration and for alternative administration such as oral, nasal, intestinal, pulmonary, blood-brain barrier, etc. Furthermore, the small size also helps reduce immunogenicity. A higher disulfide density can be obtained either by increasing the number of disulfides or by using domains with the same number of disulfides, but fewer amino acids. It is also desired to decrease the number of fixed non-cysteine residues, so that a greater percentage of amino acids are available for target binding.

[0153]Cysteine-containing binding modules can adopt a wide range of disulfide bond patterns (DBPs). For example, two-disulfide modules can have three different disulfide bond patterns (DBPs), three-disulfide modules can have 15 different DBPs, and four-disulfide modules have up to 105 different DBPs. Natural examples exist for all 2SS DBPs, most 3SS DBPs, and less than half of the 4SS DBPs.

[0154]In one aspect, the natural or non-naturally occurring cysteine (C)-containing module comprises a polypeptide having two disulfide bonds formed by pairing of cysteines contained in the polypeptide according to a pattern selected from the group consisting of C1- 2, 3-4, C1-3, 2-4 and C1-4, 2-3, in which the two numerical numbers joined by a hyphen indicate that two cysteines counting from the N terminus of the polypeptide pair to form a bond disulfide. In another aspect, the natural or non-naturally occurring cysteine (C)-containing module comprises a polypeptide having three disulfide bonds formed by pairing of cysteines between scaffolds according to a pattern selected from the group consisting of C1-2, 3-45 -6, C1-2, 3-5, 4-6, C1-2, 3-6, 4-5, C1-3, 2-4, 5-6, Cl-3, 2-5, 4-6 , C1-3, 2-6, 4-4, C1-4, 2-3, 5-6, C1-4, 2-6, 3-5, C1-5, 2-2, 4-6, C1 -5, 2-4, 3-6, C1-5, 2-6, 3-4, C1-6, 2-3, 4-5 and C1-6, 2-5, 3-4, in which The two numerical numbers joined by a hyphen indicate that two cysteines counting from the N terminus of the polypeptide pair to form a disulfide bond. In yet another aspect, the natural or non-naturally occurring cysteine (C)-containing module comprises a polypeptide having at least four disulfide bonds formed by pairing of cysteines contained in the polypeptide according to a pattern selected from the group of permutations defined by the previous formula. In yet another aspect, the natural or non-naturally occurring cysteine (C)-containing module comprises a polypeptide having at least five, six or more disulfide bonds formed by pairing of cysteines between proteins according to a pattern selected from the group of permutations represented by the previous formula. Any of the cysteine-containing proteins or scaffolds disclosed in the co-pending patent applications [serial numbers 11/528,927 and 11/528,950] are candidate binding modules.

[0155]The binding modules can also be selected from libraries of cysteine-restricted cyclic peptides with 4, 5, 6, 7, 8, 9, 10, 11 and 12 amino acids randomized or partially randomized between the disulfide-linked cysteines (e.g. , in a formation mode), and in some cases additional randomized amino acids on the outside of the cysteine pair can be constructed using a variety of procedures. Members of libraries with specificity for a target of interest can be identified using various methods including phage display, ribosomal display, yeast display, and other methods known in the art. Such cyclic peptides can be used as binding modules in MURP. In a preferred embodiment, cysteine-restricted peptides can be further manipulated to increase binding affinity, proteolytic stability and/or specificity using formation approaches that lead to binding modules containing more than one disulfide bond. A particular training approach is illustrated in Fig. 25. It is based on the addition of a single cysteine plus multiple randomized residues on the N-terminal side of the previously selected cyclic peptide, in addition to on the C-terminal side. Libraries can be generated which have been designed as illustrated in Fig. 25. Binding modules with improved properties can be identified by phage display or similar procedures. Such training libraries may contain between 1 and 12 random positions on the N-terminus as well as on the C-terminus side of a cyclic peptide. The distance between the cysteine residues on the newly added random flanks and the cysteine residues on the cyclic peptide can be varied between 1 and 12 residues. Such libraries will contain four cysteine residues per library member, with two cysteines resulting from the original cyclic peptide and two cysteine residues on the newly added flanks. This approach favors a 1-4 2-3 DBP or a change in DBP, cleaving the pre-existing 1-2 disulfide (= 2-3 in the 4-cysteine construct) to form a 1-23-4 or 1-32 4 DBP. Such training approaches can be performed with clone-specific primers so that no fixed sequence remains between areas of the library as shown in Fig. 25, or they can be performed with primers that use (and then remain) a fixed sequence in both sides of the previously selected peptide and, therefore, these same primers can be used for any previously selected clone as illustrated in Fig. 26. The procedure illustrated in Fig. 26 can be applied to a set of cyclic peptides with specificity for a target of interest. Both training approaches were shown to work for anti-VEGF affinity maturation by training. This approach can be repeated to generate binding modules with six or more cysteine residues.

[0156]Another formation of a disulfide in a 2-disulfide sequence is illustrated in Fig. 27. It involves dimerization of a preselected set of 1-disulfide peptides with itself such that the preselected set of peptides terminates at the N-terminus. , in addition to the C-terminal position. This approach favors the formation of 2-disulfide sequences that recognize two separate epitopes on a target.

[0157]Another training approach involves the addition of a (partially) randomized sequence of 6-15 residues containing two cysteines that are 4, 5, 6, 7, 8, 9, or 10 amino acids apart, with optionally additional randomized positions outside. of the linked cysteines. This random sequence of 2 cysteines is added on the N-terminal side of the previously selected peptide, or on the C-terminal side. This approach favors a 1-23-4 DBP, although other DBPs can be formed. This approach can be repeated to generate binding modules with six or more cysteine residues.

[0158]Joining modules can be built based on scaffolding of natural structures. Such scaffolds can be identified by database searching. Libraries that are based on natural scaffolds can undergo immunopurification by phage display, followed by screening to identify sequences that specifically bind to a target of interest.

[0159]A wide selection of natural scaffolds is available to construct the binding modules. The choice of a particular scaffold will depend on the intended target. Non-limiting examples of natural scaffolds include snake toxin-like proteins such as snake venom toxins and extracellular domain of human cell surface receptors. Non-limiting examples of snake venom toxins are erabutoxin B, gamma-cardiotoxin, faciculin, muscarinic toxin, erabutoxin A, neurotoxin I, cardiotoxin V4II (toxin III), cardiotoxin V, alpha-cobratoxin, long neurotoxin 1, FS2 axin, bungarotoxin , bucandin, cardiotoxin CTXI, cardiotoxin CTX IIB, cardiotoxin II, cardiotoxin III, cardiotoxin IV, cobrotoxin 2, alpha-toxins, neurotoxin II (cobrotoxin B), toxin B (neurotoxin), candotoxin, bucaine. Non-limiting examples of extracellular domain of cell surface receptors (human) include CD59, activin type II receptor, BMP receptor Ia ectodomain, TGF-beta type II receptor extracellular domain. Other natural scaffolds include, but are not limited to, A-domains, EGF, Ca-EGF, TNF-R, Notch, DSL, Trefoil, PD, TSP1,<t>S<p>2, TSP3, annatto, beta integrin , thyroglobulin, defensin 1, defensin 2, cyclotide, SHKT, disintegrins, myotoxins, gamma-thioneins, conotoxin, mu-conotoxin, omega-atrachotoxins, delta-atrachotoxins, in addition to additional families disclosed in the patent applications pending together with the present serial number 11/528,927 and 11/528,950, which are incorporated herein in their entirety.

[0160]A wide variety of procedures have been described that allow binding molecules to be identified in a large library of variants. One procedure is chemical synthesis. Library members can be synthesized on beads so that each bead carries a different peptide sequence. Beads carrying ligands with desirable specificity can be identified using labeled binding components. Another approach is the generation of peptide sub-libraries that allow specific binding sequences to be identified in an iterative procedure (Pinilla, C. et al. (1992) BioTechniques, 13: 901-905). Most commonly used are display procedures in which a library of variants is expressed on the surface of a phage, protein or cell. These procedures have in common that the DNA or RNA that encodes each variant in the library is physically linked to the ligand. This allows the ligand of interest to be detected or recovered and then its peptide sequence determined by sequencing the bound DNA or RNA. The display procedures allow one skilled in the art to enrich library members with desirable binding properties from large libraries of random variants. Frequently, variants with desirable binding properties can be identified from enriched libraries by screening individual strains isolated from an enriched library for desirable properties. Examples of visualization procedures are fusion with the lac repressor (Cull, M. et al. (1992) Proc. Natl. Acad. Sci. USA, 89: 1865-1869), cell surface visualization (Wittrup, K. D. (2001) Curr Opin Biotechnol, 12: 395-9). Of particular interest are methods in which random peptides or proteins are ligated to phage particles. Phage M13 (Smith, G. P. et al. (1997) Chem Rev, 97: 391-410) and phage T7 (Danner, S. et al. (2001) Proc Natl Acad Sci U S A, 98: 12954-9) are commonly used. . There are multiple procedures available to visualize peptides or proteins on phage M13. In many cases, the library sequence is fused to the N terminus of the phage M13 peptide pIII. The phage normally carries 3-5 copies of this protein and so the phage in such a library will in most cases carry between 3-5 copies of a member of the library. This approach is called multivalent visualization. An alternative is phagemid visualization in which the library is encoded on a phagemid. Phage particles can be formed by infection of cells carrying a phagemid with a helper phage (Lowman, H. B. et al. (1991) Biochemistry, 30: 10832-10838). This procedure usually leads to monovalent visualization. In some cases, monovalent display is preferred to obtain high affinity binders. In other cases, multivalent visualization is preferred (O'Connell, D. et al. (2002) J Mol Biol. 321: 49-56).

[0161]A variety of procedures have been described to enrich sequences with desirable characteristics by phage display. A target of interest can be immobilized by binding to immunotubes, microtiter plates, magnetic beads or other surfaces. Subsequently, a phage library is contacted with the immobilized target, phage lacking a binding ligand is washed, and phage carrying a target-specific ligand can be eluted using a variety of conditions. Elution can be performed at low pH, high pH, urea, or other conditions that tend to break protein-protein contacts. Bound phage can also be eluted by adding E. cells. colide so that the elution of the phage can directly infect the host of E. co-added. An interesting protocol is protease elution that can degrade phage-bound ligand or immobilized target. Proteases can also be used as tools to enrich for protease-resistant phage-bound ligands. For example, a library of phage-bound ligands can be incubated with one or more proteases (human or mouse) prior to immunopurification on the target of interest. This procedure degrades and removes protease-labile ligands from the library (Kristensen, P. et al. (1998) Fold Des, 3: 321-8). Phage display libraries can also be enriched for ligands to bind to complex biological samples. Examples are immunopurification on immobilized cell membrane fractions (Tur, M. K. et al. (2003) Int J Mol Med, 11: 523-7), or whole cells (Rasmussen, U. B. et al. (2002) Cancer Gene Ther, 9 : 606-12; Kelly, K. A. et al. (2003) Neoplasia, 5: 437-44). In some cases immunopurification conditions have to be optimized for the enrichment of cell-specific binders from phage libraries (Watters, J. M. et al. (1997) Immunotechnology, 3: 21-9). Phage immunopurification can also be performed in patients or live animals. This approach is of particular interest for the identification of ligands that bind to vascular targets (Arap, W. et al. (2002) Nat Med, 8: 121-7).

[0162]A variety of cloning procedures are available that allow one skilled in the art to generate libraries of DNA sequences encoding libraries of peptides. Random mixtures of nucleotides can be used to synthesize oligonucleotides containing one or multiple random positions. This procedure allows controlling the number of random positions, in addition to the degree of randomization. Furthermore, random or semi-random DNA sequences can be obtained by partial digestion of DNA from biological samples. Random oligonucleotides can be used to construct plasmid or phage libraries that are randomized to predefined locations. This can be done by fusion PCR as described in (de Kruif, J. et al. (1995) J Mol Biol, 248: 97-105). Other protocols are based on DNA ligation (Felici, F. et al. (1991) J Mol Biol, 222: 301-10; Kay, B. K. et al. (1993) Gene, 128: 59-65). Another commonly used approach is Kunkel mutagenesis in which a mutagenized strand of a plasmid or phagemid is synthesized using single-stranded cyclic DNA as a template. See, Sidhu, S. S. et al. (2000) Methods Enzymol, 328: 333-63; Kunkel, T. A. et al. (1987) Methods Enzymol, 154: 367-82.

[0163]Kunkel mutagenesis uses templates containing randomly incorporated uracil bases that can be obtained from strains of E. coli like CJ236. The uracil-containing template strand is preferentially degraded upon transformation into E. coli, while the mutagenized strand synthesized in vitro is retained. As a result, most transformed cells carry the mutagenized version of the phagemid or phage. A valuable approach to increasing diversity in a library is to combine multiple sub-libraries. These sublibraries can be generated by any of the procedures described above and can be based on the same or different scaffolds.

[0164]A useful procedure for generating large phage libraries of short peptides has recently been described (Scholle, M. D. et al. (2005) Comb Chem High Throughput Screen, 8: 545-51). This procedure is related to the Kunkel approach, but does not require the generation of single-stranded messenger DNA containing random uracil bases. Instead, the procedure begins with a template phage that carries one or more mutations close to the area to be mutagenized and said mutation renders the phage non-infectious. The procedure uses a mutagenic oligonucleotide that carries randomized codons in some positions and that corrects the fado-inactivating mutation in the template. As a result, only mutagenized phage particles are infectious after transformation and very few parental phages are contained in such libraries. This procedure can be further modified in several ways. For example, multiple mutagenic oligonucleotides can be used to simultaneously mutagenize multiple discontiguous regions of a phage. The present inventors have taken this approach a step further by applying it to complete microproteins of >25, 30, 35, 40, 45, 50, 55 and 60 amino acids, instead of to short peptides of <10, 15 or 20 amino acids, which poses an additional challenge. This approach now gives libraries of more than 10e10 transformants (up to 10e11) with a single transformation, so that from 10 transformations a single library with a diversity of 10e12 is expected.

[0165]Another variation of the Scholle procedure is to design the mutagenic oligonucleotide so that an amber stop codon on the template becomes an ocher stop codon, and an ocher stop codon becomes an amber one in the next cycle of mutagenesis. In this case, the template phage and the mutagenized library members must be grown on different suppressor strains of E. coli, alternating an ocher suppressor with amber suppressor strains. This allows successive rounds of phage mutagenesis to be carried out alternating between these two types of stop codons and two suppressor strains.

[0166]Still another variation of Scholle's approach involves the use of megaprimers with a single-stranded phage DNA template. The megaprimer is a long ssDNA that was generated from the phage curated set library inserts from the previous round of immunopurification. The goal is to capture the full diversity of library inserts from the previous set, which was mutagenized in one or more areas, and transfer it to a new library such that an additional area can be mutagenized. The megaprimer procedure can be repeated for multiple cycles using the same template containing a stop codon in the gene of interest. The megaprimer is a ssDNA (optionally generated by PCR) that contains 1) 5' and 3' overlap areas of at least 15 bases for complementarity with the ssDNA template, and 2) one or more previously selected library areas ( 1, 2, 3, 4 or more) that were copied (optionally by PCR) from the set of previously selected clones, and 3) a newly mutagenized library area that is to be selected in the next round of immunopurification. The megaprimer is optionally prepared by 1) synthesizing one or more oligonucleotides encoding the area of the newly synthesized library and 2) fusing this, optionally using overlap PCR, with a DNA fragment (optionally obtained by PCR) containing any other area of library that was previously optimized. The final round or single-stranded PCR of the combined (overlapping) PCR product is used to generate the single-stranded megaprimer containing all previously optimized areas of the new library for an area to be optimized in the next immunopurification experiment. This approach is expected to enable affinity maturation of proteins using multiple rapid cycles of library creation that generate 10e11 to 10e12 diversities per cycle, each followed by immunopurification.

[0167]A variety of procedures can be applied to introduce sequence diversity into libraries (previously selected or naïve) of microproteins or to mutate individual microprotein clones with the aim of enhancing their binding or other properties such as manufacturing, stability or immunogenicity. In principle, all procedures that can be used to generate libraries can also be used to introduce diversity into enriched (previously curated) libraries of microproteins. In particular, variants with desirable binding or other properties can be synthesized and partially randomized oligonucleotides designed based on these sequences. This procedure allows controlling the positions and degree of randomization. The utility of individual mutations in a protein can be inferred from multi-variant sequence data using a variety of computer algorithms (Jonsson, J. et al. (1993) Nucleic Acids Res. 21: 733-9; Amin. N. et al. . (2004) Protein Eng Des Sel, 17: 787-93). Of particular interest for re-mutagenesis of enriched libraries is DNA shuffling (Stemmer, W. P. C. (1994) Nature, 370: 389 391), which generates recombinants of individual sequences in an enriched library. Shuffling can be performed using a variety of modified PCR conditions and templates can be partially degraded to enhance recombination. An alternative is recombination at predefined positions using restriction enzyme-based cloning. Of particular interest are procedures that use type IIS restriction enzymes that cleave DNA outside its sequence recognition site (Collins, J. et al. (2001) J Biotechnol. 74: 317-38). Restriction enzymes that generate non-palindromic overhangs can be used to excise plasmids or other mixtures of DNA-encoding variants at multiple locations and complete plasmids can be reassembled by ligation (Berger, S. L. et al. (1993) Anal Biochem, 214: 571-9 ). Another procedure for introducing diversity is PCR mutagenesis in which DNA sequences encoding members of the libraries are subjected to PCR under mutagenic conditions. PCR conditions have been reported to lead to mutations at relatively high mutation frequencies (Leung, D. et al. (1989) Technique, 1: 11-15). Furthermore, a polymerase with reduced fidelity can be used (Vanhercke, T. et al. (2005) Anal Biochem, 339: 9-14). A procedure of particular interest is based on mutator strains (Irving, R. A. et al. (1996) Immunotechnology, 2: 127-43; Coia, G. et al. (1997) Gene, 201: 203-9). These are strains that carry defects in one or more DNA repair genes. The plasmids or phage or other DNA in these strains accumulate mutations during normal replication. Individual clones or enriched populations can be propagated in mutator strains to introduce genetic diversity. Many of the procedures described above can be used in an iterative procedure. Multiple rounds of mutagenesis and screening or immunopurification can be applied to entire genes, or to portions of a gene, or different portions of a protein can be mutagenized during each subsequent round (Yang, W. P. et al. (1995) J Mol Biol, 254: 392- 403).

[0168]Libraries may be additionally treated to reduce artifacts. Known artifacts of phage immunopurification include 1) nonspecific binding based on hydrophobicity, and 2) multivalent binding to the target, both due to a) the pentavalence of the phagopIII protein and b) due to the formation of disulfides between different microproteins, producing multimers. , or c) due to high-density coating of the target on a solid support and 3) context-dependent target binding, in which the context of the target or the context of the microproteins becomes critical for the binding or binding activity. inhibition. Different stages of treatment can be performed to minimize the magnitude of these problems. For example, such treatments are applied to the entire library, but some useful treatments that remove bad clones can only be applied to sets of soluble proteins or only to individual soluble proteins.

[0169]Cysteine-containing scaffold libraries are likely to contain free thiols, which may complicate evolution directed by cross-linking to other proteins. One approach is to remove the worst clones from the library by running them through a thiol-free column, thus eliminating all clones that have one more free sulfhydryls. Clones with free SH groups can also be reacted with biotin-SH reagents, allowing efficient removal of clones with reactive SH groups using streptavidin columns. Another approach is not to remove free thiols, but to inactivate them by end capping with sulfhydryl-reactive chemicals such as iodoacetic acid. Of particular interest are bulky or hydrophilic sulfhydryl reagents that reduce non-specific target binding or modified variants.

[0170]Examples of context dependence are all constant sequences, including the pIII protein, linkers, peptide tags, biotin-streptavidin, Fc and other fusion proteins that contribute to the interaction. The typical approach

as practical in order to avoid formation. This may involve alternating between different display systems (i.e., M13 vs. T7, or M13 vs. yeast), alternating the labels and linkers used, alternating the (solid) support used for immobilization (i.e., immobilization) and alternate the target proteins themselves (different vendors, different fusion versions).

[0171]Library screening can also be used to select proteins with preferred qualities. One option is the treatment of libraries with proteases in order to eliminate unstable variants from the library. The proteases used are normally those that would be found in the application. For pulmonary administration, pulmonary proteases would be used, for example, obtained by lung lavage. Similarly, mixtures of proteases would be obtained from serum, saliva, stomach, intestine, skin, nose, etc. However, it is also possible to use mixtures of individual purified proteases. A comprehensive list of proteases is shown in [Appendix E]. Phages themselves are exceptionally resistant to most proteases and other harsh treatments.

[0172]For example, it is possible to select the library for the most stable structures, that is, those with the strongest disulfide bonds, by exposure to increasing concentrations of reducing agents (that is, DTT or beta-mercaptoethanol), thus first removing less stable structures. Typically reducing agent (i.e. DTT, BME, other) would be used at concentrations of 2.5mM, 5mM, 10mM, 20mM, 30mM, 40mM, 50mM, 60mM, 70mM, 80mM , 90 mM or even 100 mM, depending on the desired stability.

[0173]It is also possible to select clones that can refold efficiently in vitro, by reducing the entire display library with a high level of reducing agent, followed by gradually reoxidizing the protein library to reform the disulfides, followed by elimination of clones with free SH groups, as described above. This procedure can be applied once or multiple times to eliminate clones that have low in vitro refolding efficiency.

[0174]One approach is to apply genetic selection for the level of protein expression, folding and solubility as described by A. C. Fisher et al. (2006) Genetic selection for protein solubility enabled by the folding quality control feature of the twin-arginine translocalization pathway. Protein Science (online). After immunopurification of display libraries (optional), one would want to avoid screening thousands of clones at the protein level for target binding, expression level and folding. An alternative is to clone the complete set of selected inserts into a beta-lactamase fusion vector which, when plated on beta-lactam, the authors showed was selective for well-expressed, fully disulfide-linked, and soluble proteins.

[0175]Following M13 phage display of protein libraries and immunopurification on targets for one or more cycles, there are a variety of ways to proceed, including (1) screening individual phage clones by phage ELISA, which measures the number of phage particles (using anti-M13 antibodies) that bind to an immobilized target; (2) transfer phage display libraries from M13 to T7. The second approach is particularly useful in reducing the occurrence of false positives based on valence. Any single library format tends to favor clones that can form high avidity contact with the target. This is why screening for soluble proteins is important, although this is a tedious solution. The multivalency achieved in T7 phage display is likely very different from that achieved in M13 display, and cycling between T7 and M13 may be an excellent approach to reduce the occurrence of false positives based on valence.

[0176]Filter elevation is another methodology that can be with bacterial colonies grown at high density on large agar plates (10e2-10e5). Small amounts of some proteins are secreted into the media and end up attached to the filtration membrane (nitrocellulose or nylon). The filters are then blocked in skimmed milk, 1% casein hydrolyzate, or a 1% BSA solution and incubated with the target protein that has been labeled with a fluorescent dye or reporter enzyme (directly or indirectly by antibodies or through biotin-streptavidin). Colony location is determined by coating the filter on the bottom of the plate and all positive colonies are selected and used for further characterization. The advantage of filter elevations is that they can be made affinity selective by reading the signal after washing for different periods of time. The signal from high-affinity clones 'fades' slowly, while the signal from low-affinity clones 'fades' rapidly. Such affinity characterization typically requires a 4-point assay with a substantiated assay and may provide better clone-to-clone comparability than substantiated assays. Milling colonies on a matrix is useful as it minimizes differences due to colony size or location.

N-end modules:

[0177]The subject MURPs may contain N-end modules (NM), which are particularly useful, for example, in facilitating the production of the MURPs. The NM may be a single methionine residue if the products are expressed in the cytoplasm of E. coli. A typical product format is a URP fused to a therapeutic protein, which is expressed in the bacterial cytoplasm so that the N terminus is formyl-methionine. Formylmethionine can be either permanent or temporary, if removed by biological or chemical processing.

[0178]The NM can also be a peptide sequence that has been manipulated for proteolytic processing, which can be used to remove tags or to remove fusion proteins. The N-terminal module can be manipulated to facilitate MURP purification, including an affinity tag such as the Flag, Myc, HA or His tag. The N-terminal module may also include an affinity tag that may be used for MURP detection. An NM can be manipulated or selected for high-level expression of the MUMP. It can also be manipulated or selected to enhance the protease resistance of the resulting MURP. MURPs can be produced with an N-terminal module that facilitates expression and/or purification. This N-terminal module can be cleaved during the production process with a protease, so that the final product does not contain an N-terminal module.

[0179]Optimizing the choice of amino acids and codons of the N-terminal module can increase recombinant production. The N-terminal module may also contain a processing site that can be cleaved by a specific protease such as factor Xa, thrombin, or enterokinase. Tomato etch virus protease (vGt). Processing sites can also be designed to be cleavable by chemical hydrolysis. An example is the asp-pro amino acid sequence which can be cleaved under acidic conditions. An N-terminal module can also be designed to facilitate purification of a MURP. For example, N-terminal modules can be designed to contain multiple his residues that allow product capture by immobilized metal chromatography. The N-terminal modules may contain peptide sequences that can be specifically captured or detected by antibodies. Examples are FLAG, HA, c-myc.

C-end modules:

[0180]MURPs may contain a C-end module, which are particularly useful, for example, in facilitating the production of MURPs. For example, the C-terminal module may comprise a cleavage site to effect proteolytic processing to remove sequences that are fused and thereby increase protein expression or facilitate purification. In particular, the C-terminal module may also contain a processing site that can be cleaved by a specific protease such as factor Xa, thrombin, VGT protease or enterokinase. Processing sites can also be designed to be cleavable by chemical hydrolysis. An example is the asp-pro amino acid sequence which can be cleaved under acidic conditions. The C-terminal module may be an affinity mark that aims to facilitate MURP purification. For example, C-terminal modules can be designed to contain multiple his residues that allow product capture by immobilized metal chromatography. The C-terminal modules may contain peptide sequences that can be specifically captured or detected by antibodies. Non-limiting examples of the tags include FLAG, HA, c-myc or His tag. The C-terminal module can also be manipulated or selected to enhance the protease resistance of the resulting MURP.

[0181]If desired, the N-terminus of the protein can be ligated to its own C-terminus. For example, ligation of these two modules can be carried out by creating a natural amino acid-like bond (peptide bond) or using an entity of exogenous link. Of particular interest are cyclotides, a family of small proteins in which this occurs naturally. Adopting a structural format such as cyclotides is expected to provide additional stability against exo-proteases. Such intramolecular bonding typically works better at lower protein concentrations.

Effector modules:

[0182]MURPs may comprise one or multiple effector modules (EMs), or none. Effector modules typically do not provide targeting, but provide an activity required for therapeutic effect, such as cell destruction. EMs can be small pharmaceutically active molecules (i.e., toxic drugs), peptides, or proteins. Non-limiting examples are cytokines, antibodies, enzymes, growth factors, hormones, receptors, receptor agonists or antagonists, whether complete or a fragment or domain thereof. Effector modules may also comprise peptide sequences carrying chemically linked small molecule drugs, whether synthetic or natural. Optionally, these effector molecules may be linked to the effector module by chemical linkers, which may or may not cleave under selected conditions leading to a release of toxic activity. EMs may also include radioisotopes and their chelates, as well as various labels for PET and MRI. Effector modules can also be toxic to a cell or tissue. Of particular interest are MURPs that contain toxic effector modules and binding modules with specificity for a diseased tissue or diseased cell type. Such MURPs can accumulate specifically in a diseased tissue or in diseased cells and can exert their toxic action preferentially in the diseased cells or tissues. Example effector modules are listed below.

[0183]Enzymes -The effector modules can be enzymes. Of particular interest are enzymes that degrade metabolites that are critical for cell growth such as carbohydrates or amino acids or lipids or co-factors. Other examples of effector modules with enzymatic activity are RNase, DNase and phosphatase, asparaginase, histidinase, arginase, beta-lactamase. Effector modules with enzymatic activity can be toxic when administered to a tissue or cell. Of particular interest are MURPs that combine effector modules that are toxic and binding modules that bind specifically to diseased tissue. Also possible effector modules are enzymes that convert an inactive prodrug into an active drug at the tumor site.

[0184]Drug -The MURP object may contain an effector that is a drug. If desired, sequences can be designed for organ-selective delivery of drug molecules. An example is illustrated in Figure 8. A URP sequence can be fused to a protein that preferentially binds to diseased tissue. The same URP sequence may contain one or more amino acid residues that can be modified for the binding of drug molecules. Such a conjugate can bind to diseased tissue with high specificity and the bound drug molecules can produce local action, while maintaining systemic drug exposure. MURP can be designed to facilitate the release of drug molecules at the target size by introducing protease-sensitive sites that can be cleaved by native proteases at the site of desired action. A significant advantage of using URP sequences for the design of drug delivery constructs is that undesirable interactions between the drug molecule and the targeting domain of the construct can be avoided. Many drug molecules that can be conjugated to targeting domains have significant hydrophobicity and the resulting conjugates tend to aggregate. Adding hydrophilic URP sequences to such constructs can improve the solubility of the resulting delivery constructs and, consequently, reduce the tendency for aggregation. Furthermore, the number of drug molecules that can be fused to a targeting domain can be increased by adding long URP sequences. Furthermore, the use of URP sequences allows the distance between drug conjugation sites to be optimized to facilitate complete conjugation. The list of suitable drugs includes, but is not limited to, chemotherapeutic agents such as thiotepa and cyclophosphamide (CYTOXAN™); alkylsulfonates such as busulfan, improsulfan and piposulfan; aziridines such as benzodopa, carboquone, meturedopa, and uredopa; ethyleneimines and methylmelamines including altretamine, triethylenemelamine, triethylenephosphoramide, triethylenethiophosphoramide and trimethylololamine; nitrogen mustards such as chlorambucil, chlornaphazine, cyclophosphamide, estramustine, ifosfamide, mechlorethamine, mechlorethamine oxide hydrochloride, melphalan, novembiquin, phenesterin, prednimustine, trofosfamide, uracil mustard; nitrosoureas such as carmustine, chlorozotocin, fotemustine, lomustine, nimustine, ranimustine; antibiotics such as aclacinomycins, actinomycin, autramycin, azaserine, bleomycins, cactinomycin, carabicin, carminomycin, carzinophilin, chromomycins, dactinomycin, daunorubicin, detorubicin, 6-diazo-5-oxo-L-norleucine, doxorubicin, epirubicin, esorubicin, idarubicin, marcelomycin , mitomycins, mycophenolic acid, nogalamycin, olivomycins, peplomycin, potfiromycin, puromycin, chelamycin, rhodorubicin, streptonigrin, streptozocin, tubercidin, ubenimex, zinostatin, zorubicin; antimetabolites such as methotrexate and 5-fluorouracil (5-FU); folic acid analogues such as denopterin, methotrexate, pteropterin, trimetrexate; purine analogues such as fludarabine, 6-mercaptopurine, thiamiprine, thioguanine; pyrimidine analogues such as ancitabine, azacitidine, 6-azauridine, carmofur, cytarabine, dideoxyuridine, doxyfluridine, enocitabine, floxuridine, androgens such as calusterone, dromostanolone propionate, epithiostanol, mepitiostane, testolactone; antiadrenergics such as aminoglutethimide, mitotane, trilostane; folic acid repairers such as frolinic acid; aceglatone; aldophosphamide glucoside; aminolevulinic acid; amsacrine; bestrabucil; bisantrene; edatraxate; defofamine; demecolcine; diaziquone; duocarmycin, maytansine, auristatin, elfornithine; elliptinium acetate; ethoglucide, gallium nitrate; hydroxyurea; lentinan; lonidamine; mitoguazone; mitoxantrone; mopidamol; nitracrine; pentostatin; fenamet; pyrarubicin; podophylinic acid; 2-ethylhydrazide; procarbazine; PSK.R™; razoxane; sizophyran; spirogermanium; tenuazonic acid; triaziquone; 2,2',2"-trichlorotriethylamine; urethane; vindesine; dacarbazine; mannomustine; mitobronitol; mitolactol; pipobromane; gacytosine; arabinoside (“Ara-C”); cyclophosphamide; thiotepa; taxanes, e.g. paclitaxel (TAXOL™, Bristol -Myers Squibb Oncology, Princeton, N.J.) and doxetaxel (TAXOTERE™, Rhóne-Poulenc Rorer, Antony, France); etoposide (VP-16); DMFO); retinoic acid; Esperamycins; and pharmaceutically acceptable salts, acids or derivatives of any of the above. Also included are antihormonal agents that act to regulate or inhibit the action of hormones such as antiestrogens. include, for example, tamoxifen, raloxifene, 4(5)-imidazoles aromatase inhibitors, 4-hydroxytamoxifen, trioxifene, keoxifene, LY117018, onapristone and toremifene (Fareston); and antiandrogens such as flutamide, nilutamide, bicalutamide, leuprolide, goserelin, doxorubicin, daunomycin, duocarmycin, vincristine and vinblastine.

[0185]Other drugs that can be used as effector modules include those that are useful for treating inflammatory conditions, heart diseases, infectious diseases, respiratory diseases, autoimmune diseases, neuronal and muscular disorders, metabolic disorders and cancers.

[0186]Additional drugs that can be used as effectors in MURP include pain and inflammation agents such as histamine and histamine antagonists, bradykinin and bradykinin antagonists, 5-hydroxytryptamine (serotonin), lipid substances that are generated by biotransformation of the products of selective hydrolysis of membrane phospholipids, eicosanoids, prostaglandins, thromboxanes, leukotrienes, aspirin, non-steroidal anti-inflammatory agents, analgesic-antipyretic agents, agents that inhibit the synthesis of prostaglandins and thromboxanes, selective inhibitors of inducible cyclooxygenase, selective inhibitors of inducible cyclooxygenase-2, autacoids, paracrine hormones, somatostatin, gastrin, cytokines mediating interactions involved in humoral and cellular immune responses, peptide-derived autacoids, eicosanoids, p-adrenergic agonists, ipratropium, glucocorticoids, methylxanthines, channel blockers sodium, opioid receptor agonists, calcium channel blockers, membrane stabilizers and leukotriene inhibitors.

[0187]Other drugs that can be used as an effector include agents for the treatment of peptide ulcers, agents for the treatment of gastroesophageal reflux disease, prokinetic agents, antiemetics, agents used in irritable bowel syndrome, agents used for diarrhea, agents used for constipation, agents used for inflammatory bowel disease, agents used for biliary disease, agents used for pancreatic disease.

[0188]Radionuclides -MURPs can be designed for tissue-targeting delivery of radionuclides, as well as for radionuclide imaging. URPs are ideal for imaging because the half-life can be optimized by changing the length of the URP. For most imaging applications a moderately long URP is likely to be preferred, providing a half-life of 5 minutes to a few hours, not days or weeks. MURPs can be designed to contain only a single amino group or a small defined number of amino groups that can be modified with chelating agents (such as DOTA) for radioisotopes such as technetium, indium and yttrium (EXPAND). Alternative conjugation procedures are via reserved cysteine side chains. Such radionuclide-bearing MURPs can be used for the treatment of tumors or other diseased tissues, as well as for imaging.

[0189]Many pharmaceutically active proteins or protein domains can be used as effector models in MURP. Examples are the following proteins, as well as fragments of these proteins: cytokines, growth factors, enzymes, receptors, microproteins, hormones, erythropoietin, adenosine deiminase, asparaginase, arginase, interferon, growth hormone, growth hormone-releasing hormone, G -CSF, GM-CSM, insulin, hirudin, TNF receptor, uricase, rasburicase, axokine, RNase, DNAse, phosphatase, Pseudomonas exotoxin, ricin, gelonin, desmoteplase, laronidase, thrombin, blood clotting enzyme, VEGF, protropin , somatropin, alteplase, interleukin, factor IIV, factor VIII, factor

[0190] Protease-activated MURP: To enhance the therapeutic index of an effector module, protease-labile sequences can be inserted into URP sequences that are sensitive to proteases that are preferentially found in serum or in the target tissue to be treated by MURP. . This approach is illustrated in Figure 9. Some designs allow the construction of proteins that are selectively activated when they reach a target tissue. Of particular interest are MURPs that are activated at a disease site. To facilitate such target-specific activation, URP sequences can be linked in close proximity to the active site or receptor binding site of the effector module so that the resulting fusion protein has limited biological activity. Of particular interest is the activation of an effector module at a tumor site. Many tumor tissues express proteases at relatively high concentrations and sequences that are specifically cleaved by these tumor proteases can be inserted into URP sequences. For example, most prostate tumor tissues contain high concentrations of prostate-specific antigen (PSA), which is a serine protease. Prodrugs consisting of a PSA-labile peptide conjugated to the anticancer drug doxorubicin have shown selective activation in prostate tissue [DeFeo-Jones, D. et al. (2000) Nat Med, 6: 1248]. Of particular interest for disease-specific activation are proteins with cytostatic or cytotoxic activity such as TNF-alpha, and many cytokines and interleukins. Another application is the selective activation of proteins at the site of inflammation or at the site of virus or bacterial infection.

[0191]Production procedures -MURPs containing URP sequences can be produced using molecular biology approaches that are well known in the art. A variety of cloning vectors are available for various expression systems such as mammalian cells, yeast, and microbes. Of particular interest as expression hosts are E. coli, S. cerevisiae, P. pastoris, and Chinese hamster ovary cells. Of particular interest are guests that have been optimized to expand their codon usage. Of particular interest is a host that has been modified to enhance GRS expression. This can be done by providing DNA that encodes glycine-specific tRNAs. Furthermore, the host can be manipulated in a way that enhances the loading of glycine-specific tRNA. The DNA encoding the enhanced protein can be operationally linked to promoter sequences. The DNA encoding the enhanced protein, in addition to the operationally linked promoter, may be part of a plasmid vector, viral vector, or may be inserted into the host chromosome.

[0192] For production, the host can be grown under conditions that facilitate production of the enhanced protein. Of particular interest are conditions that enhance LFG production.

[0193] Object MURPs can take a variety of formats. For example, MURPs may contain URPs that are fused to pharmaceutically active proteins to produce slow-release products. Such products may be injected or implanted locally, for example, in or under the skin of a patient. Due to its large hydrodynamic radius, the product containing URP sequences is slowly released from the injection or implantation site leading to a reduction in the injection or implantation frequency. URP sequences can be designed to contain regions that bind to cell surfaces or tissue in order to prolong local retention of the drug at the injection site. Of particular interest are products containing URPs that may be formulated as soluble compounds, but form aggregates or precipitates upon injection. This aggregation or precipitation can be triggered by a change in pH between the formulated product and the pH at the injection site. Alternatives are URP-containing products that precipitate or form aggregates as a result of a change in redox conditions. Still another approach is a URP-containing product that is stabilized in solution by the addition of non-active solutes, but that precipitates or aggregates upon injection as a result of diffusion of the solubilizing solutes. Another approach is to design URP-containing products that contain one or multiple lysine or cysteine residues in their URP sequence and that can be cross-linked prior to injection.

[0194] If desired, MURP is monomeric (here meaning non-cross-linked) when manufactured and formulated and when injected, but after subcutaneous injection the protein begins to cross-link with itself or with native human proteins, forming a polymer under the skin from which active drug molecules are released only very gradually. Such release may be by disulfide bond reduction or disulfide shuffling as illustrated in Fig. 18, or may be mediated by proteolysis as shown in Fig. 19, releasing active fragments into the circulation. It is important that these active fragments are large enough to have a long half-life, because the longer their secretion half-life, the lower the dose of the protein released, allowing the use of a lower dose of product to be injected. or a long time between injections.

[0195]One approach that offers these advantages is protein disulfide-mediated cross-linking. For example, a protein drug would be made with a cyclic peptide in it (one or more). This cyclic peptide may or may not participate in binding to the target. This protein is manufactured with the formed cyclic peptide, that is, in oxidized form, to simplify purification. However, the product is then reduced and formulated to keep the protein in a reduced form. It is important that the cyclic peptide be reduced to a low concentration of reducing agent, such as 0.25, 0.5, 1.0, 2.0, 4.0 or 8.0 mM dithiothreitol or cysteine or equivalent reducing agent, so that the cyclic peptide can be reduced without reducing other disulfide-containing protein modules in the product. The use of FDA-approved reducing agents, such as cysteine or glutathione, is preferred. After subcutaneous injection, the low molecular weight reducing agent rapidly diffuses or is neutralized by human proteins, exposing the drug to an oxidizing environment, while at high molar concentration, producing cross-linking of cysteines located on different protein chains. , which lead to polymerization of the drug at the injection site. The greater the distance between the cysteines in the cyclic peptide, and the greater the drug concentration, the greater the degree of polymerization of the drug, since the polymerization competes with the re-formation of the cyclic peptide. Over time, the reduction and oxidation of the disulfide will produce re-shuffling of disulfide, which will lead to the re-formation of cyclic peptide and monomerization and re-solubilization of the drug. Drug release from the polymer can also occur by proteolysis, which could be targeted and controlled or increased by forming cleavage sites for serum proteases. Cross-linking of proteins could also be performed with a protein-protein chemical cross-linking agent such as those listed in [Table x]. Ideally, this is an agent already approved by the FDA, such as those used for vaccine conjugation or protein conjugation of chemicals.

[0196]Instead of using disulfides, proteins can also be stabilized against proteolytic degradation using a wide variety of cross-linking agents. Most of the following agents are marketed by Pierce Chemicals under the same name and instructions for their use are available online (www.piercenet.com). Agents that produce the same chain-to-chain distance as has been obtained with disulfides are those most likely to be useful for the present application. Short linker agents such as DFDNB are the most promising. The distance between chains can be easily determined from the structures of chemicals as shown at www.piercenet.com.

[0197]There are a large number of specific chemicals that work based on the following small number of basic reaction schemes, all of which are described in detail at www.piercenet.com. Examples of useful cross-linking agents are imidoesters, active halogens, maleimide, pyridyl disulfide, NHS ester. Homobifunctional cross-linking agents have two identical reactive groups and are frequently used in a one-step chemical cross-linking process. Examples are BS3 (a non-cleavable water-soluble DSS analogue), BSOCOES (base-reversible), DMA (dimethyl adipimidate-2HCl), DMP (dimethyl pimelimidate-2HCl), DEM (dimethyl suberimidate-2HCl), DSG (5-carbon analogue of D<s>S), DSP (Lomant reagent), DSS (non-cleavable), DST (cleavable by oxidation agents), DTBP (dimethyl 3,3'-dithiobispropionimidate-2HCl), DTSSP , EGS, Sulfo-EGS, THPP, TSAT, DFDNB (1,5-difluoro-2,4-dinitrobenzene) is especially useful for cross-linking over small spatial distances (Kornblatt, J.A. and Lake, D.F. (1980). Cross-linking of cytochrome oxidase subunits with difluorodinitrobenzene. Can J. Biochem. 58, 219-224).

[0198]Sulfhydryl-reactive homofunctional cross-linking agents that are sulfhydryl-reactive homobifunctional protein cross-linkers are frequently based on maleimides, which react with -SH groups at pH 6.5-7.5, forming stable thioether bonds. BM[PEO]3 is an 8-atom polyether spacer that reduces the chances of conjugate precipitation in sulfhydryl-sulfhydryl cross-linking applications. BM[PEO]4 is similar, but with an 11-atom spacer. BMB is a non-cleavable cross-linker with a four-carbon spacer. BMDB forms a bond that can be cleaved by periodate. BMH is a widely used homobifunctional sulfhydryl-reactive cross-linker. BMOE has an especially short linker. DPDPB and DTME are cleavable cross-linkers. HVBS does not have the hydrolysis potential of meleimides. TMEA is another option. Heterobifunctional cross-linking agents have two different reactive groups. Examples are NHS esters and amines/hydrazines by activation of EDC, AEDP, ASB<a>(photoreactive, iodable), EDC (water-soluble carbodiimide). Amine-sulfhydryl reactive bifunctional crosslinkers are AMAS, APDP, BMPS, EMCA, EMCS, GMBS, KMUA, LC-SMCC, LC-SPDP, MBS, SBAP, SIA (extra-short), SIAB, SMCC, SMPB, SMPH, SMPT, SPDP, Sulfo-EMCS, Sulfo-GMBS, Sulfo-KMUS, Sulfo-LC-SMPT, Sulfo-LC-SPDP, Sulfo-MBS, Sulfo-SIAB, Sulfo-SMCC, Sulfo-SMPB. Heterobifunctional cross-linking agents reactive with amino groups are ANB-NOS, MSA, NHS-ASA, SADP, SAED, SAND, SANPAH, SASD, SFAD, Sulfo-HSAB, Sulfo-NHS-LC-ASA, Sulfo-SADP, Sulfo-SANPAH , TFCS.

[0199]The drug tagged with a His6 tag has a different slow-release format, which is mixed and co-injected with nickel-nitrilotriacetic acid conjugated beads (Ni-NTA beads), a GMO version of those available by Qiagen. The drug would be slowly released from the beads, providing controlled and slow release as illustrated in Fig. 20. The beads are optional and can be replaced by a cross-linked polymeric nickel-nitrilotriacetic acid leading to assembly of an even larger polymer.

[0200]URP sequences may contain sequences that are known to form multimers such as alpha2D [Hill, R, et al. (1998) J Am Chem Soc, 120: 1138-1145] which was used to dimerize an antibody fragment [Kubetzko, S. et al. (2005) Mol Pharmacol, 68: 1439-54]. Examples of a useful homodimerization peptide is the SKVILFE sequence. An example of useful heterodimerization sequences is the ARARAR peptide which can form dimers with the DADADA sequence and related sequences. Multimerization can improve the biological function of a molecule by increasing its avidity and can influence the pharmacokinetic properties and tissue distribution of the resulting MURPs.

[0201]The “multimerization modules” are amino acid sequences that facilitate the formation of MURP dimers or multimers. Multimerization modules can join themselves to form dimers or multimers. Alternatively, multimerization modules can be linked to other MURP modules. These can be leucine zippers or small peptides such as hydra head activator derivatives (similar to SKVILF) that form antiparallel homopolymers, or peptides such as RARARA and DADADA, which form high affinity antiparallel heteropolymers. Using one, two or more copies of these peptides can force the formation of protein dimers, linear multimers or branched multimers.

[0202]The affinity of the association can be made by changing the type, length and composition of the peptides. Some applications require peptides that form homodimers as illustrated in Fig. 21. Other applications require heterodimers. In some cases, once associated, the peptides can lock in situ by forming disulfide bonds between the two protein chains, usually on either side of the peptides. The multimerization modules are useful for linking two MURP molecules together (head to tail, head to head, or tail to tail) as illustrated in Fig. 21. The multimerization modules can be located at either the N or C terminus with in order to form dimers. If multimerization modules are present at both ends, long linear multimers will form. If more than two multimerization modules are present per protein, branched polymer networks can form. The concepts of multimerization and chemical conjugation can be combined leading to useful half-life extension and depot formation, leading to slow release of the active drug from the depot or injection site as illustrated in Fig. 23.

[0203]Object MURPs may incorporate a generic or universal URP. One approach is to express a URP that contains a long URP module, which provides half-life and contains multiple (typically 4-10) lysines (or other sites) that allow site-specific conjugation of peptides (i.e., linear, cyclic, 2SS , 3SS, etc.) that bind to a specific target. The advantage of this approach is that the URP module is generic and can be conjugated to any target-specific peptide. Ideally, the binding of the target-specific peptide to the URP is a directed binding, such that residues on the URP can only react with one residue on the target-specific peptide and exhaustive coupling can only produce a single species, which is , for example, a URP that is linked to a peptide at each lysine. This complex behaves like a high-avidity multimer in its binding properties, but is simple to manufacture. This approach is illustrated in Fig. 24.

[0204]Subject MURPs may also incorporate URPs to effect delivery across tissue barriers. URPs can be manipulated to enhance delivery across dermal, oral, oral, intestinal, nasal, blood-brain, pulmonary, thecal, peritoneal, rectal, vaginal, or many other tissue barriers.

[0205]One of the key obstacles to oral protein administration is the sensitivity of most proteins to proteases in the digestive system. Conjugation with URP sequences can improve the protease resistance of pharmaceutically active proteins and thus facilitate their uptake. It has been shown that protein uptake in the digestive system can be improved by adding molecular carriers. The main function of these vehicles is an improvement in membrane permeability [Stoll, B. R. et al. (2000) J Control Release, 64: 217-28]. Thus, sequences can be included in URP sequences that improve membrane permeability. Many sequences that improve membrane permeability are known and examples are arginine-rich sequences [Takenobu, T. et al. (2002) Mol Cancer The, 1: 1043-9]. Thus, URP sequences can be designed that improve cellular or oral uptake of proteins by combining two functions, a reduction in proteolytic degradation of the protein of interest, as well as an increase in the membrane permeability of the fusion product. Optionally, sequence may be added to the URP sequence that is sensitive to a protease that is preferentially located in the target tissue for the drug of interest, but is stable to proteases in the digestive tract. Examples of such URP sequences are sequences that contain long GRS regions, as well as sequences that are rich in basic amino acids, in particular arginine, and facilitate membrane transfer. URP can be used in a similar way to improve protein uptake via intranasal, intrapulmonary, or other routes of administration.

Examples of specific products:

[0206]DR4/DR5 agonist -DR4 and DR5 are death receptors that are expressed on many tumor cells. These receptors can be triggered by trimerization, which leads to cell death and tumor regression. Binding domains with specificity for DR4 or DR5 can be obtained by phage immunopurification or other display procedures: These specific DR4 or DR5 binding domains can be multimerized using URP modules as linkers as illustrated in Figure 12. They are of particular MURPs of interest containing three or more binding modules with specificity for DR4 or DR5 or both. As illustrated in Figure 12, MURPs may contain additional binding modules with specificity for tumor antigens that are overexpressed in tumor tissues. This allows the construction of MURPs that accumulate specifically in tumor tissue and trigger cell death. MURPs can contain modules that bind to either DR4 or DR5. Of particular interest are MURPs that contain binding modules that bind both DR4 and DR5.

[0207]Interleukin 2 chosen as a target by tumor -Interleukin 2 (IL2) is a cytokine that can enhance the immune response to tumor tissue. However, systemic IL2 therapy is characterized by significant side effects. MURPs can be constructed in a way that combines binding domains with specificity for tumor antigens and IL2 as an effector module as illustrated in Figure 13. Such MURPs can selectively accumulate in tumor tissue and thus provoke a tumor-selective immune response, to which while minimizing the systemic side effects of cytokine therapy. Such MURPs can target a variety of tumor antigens such as EpCAM, Her2, CEA, EGFR, Thomsen-Friedenreich antigen. Of particular use are MURPs that bind to tumor antigens that show slow internalization. Similar MURPs can be designed using other cytokines or tumor necrosis factor-alpha as effector modules.

[0208]Tumor-selective asparaginase -Asparaginase is used to treat patients with acute leukemia. Both the asparaginase of E. coli and Erwinia asparaginase are used for treatment. Both enzymes can lead to immunogenicity and hypersensitive reactions. Oncaspar is the PEGylated version of asparaginase that has reduced immunogenicity. However, the protein is difficult to manufacture and is administered as a mixture of isomers. The addition of URP sequences to the ends and/or to internal loops allows direct recombinant manufacturing of an asparaginase variant that is homogeneous and has low immunogenicity. Various URP sequences and binding sites can be compared to determine the optimal position for URP sequence binding. Several other enzymes that can degrade amino acids have reported antitumor activity. Examples are arginase, methioninase, phenylalanine ammonia lyase and tryptophanase. Of particular interest is the phenylalanine ammonia lyase from Streptomyces maritimus, which has a high specific activity and does not require a cofactor [Calabrese, J. C. et al. (2004) Biochemistry, 43: 11403-16]. Most of these enzymes are of bacterial or other non-human origin and are likely to cause immune reactions. The immunogenicity of these enzymes can be reduced by adding one or more URP sequences. Furthermore, the therapeutic index and PK properties of these enzymes can be improved by increasing their hydrodynamic radius as a result of binding to URP sequences.

[0209] The target MURPs can be designed to target any cellular protein. A non-limiting list is provided below.

[0210] VEGF, VEGF-R1, VEGF-R2, VEGF-R3, Her-1, Her-2, Her-3, EGF-1, EGF-2, EGF-3, Alfa3, cMet, ICOS, CD40L, LFA -1, c-Met, ICOS, LFA-1, IL-6, B7.1, B7.2, OX40, IL-lb, TACI, IgE, BAFF or BLys, TPO-R, CD19, CD20, CD22, CD33 , CD28, IL-1-R1, TNF<a>, TRAIL-R1, complement receptor 1, FGFa, osteopontin, vitronectin, ephrin A1-A5, ephrin B1-B3, alpha-2-macroglobulin, CCL1, CCL2, CCL3 , CCL4, CCL5, CCL6, CCL7, CXCL8, CXCL9, CXCL10, CXCL11, CXCL12, CCL13, CCL14, CCL15, CXCL16, CCL16, CCL17, CCL18, CCL19, CCL20, CCL21, CCL22, PDGF, TGFb, GMCSF, SCF, p40 (IL12/IL23), IL1b, IL1 a, IL1 ra, IL2, IL3, IL4, IL5, IL6, IL8, IL10, IL12, IL15, IL23, Fas, FasL, Flt3 ligand, 41BB, ACE, ACE-2, KGF , FGF-7, SCF, netrin 1, 2, IFNa, b, g, caspase 2, 3, 7, 8, 10, ADAM S1, S5, 8, 9, 15, TS1, TS5; adiponectin, ALCAM, ALK-1, APRIL, annexin V, angiogenin, amphiregulin, angiopoietin 1, 2, 4, B7-1/CD80, B7-2/CD86, B7-H1, B7-H2, B7-H3, Bc1- 2, BACE-1, BAK, BCAM, BDNF, bNGF, bECGF, BMP2, 3, 4, 5, 6, 7, 8; CRP, cadherin 6, 8, 11; cathepsin A, B, C, D, E, L, S, V, X; CD11a/LFA-1, LFA-3, GP2b3a, GH receptor, RSV F protein, IL-23 (p40, p19), IL-12, CD80, CD86, CD28, CTLA-4, «4p1, «4p7, TNF /lymphotoxin, IgE, CD3, CD20, IL-6, IL-6R, BLYS/BAFF, IL-2R, HER2, EGFR, CD33, CD52, digoxin, Rho (D), chickenpox, hepatitis, CMV, tetanus, vaccinia, antidote, bolutinic, Trail-R1, Trail-R2, cMet, TNF-R family, such as LA NGF-R, CD27, CD30, CD40, CD95, lymphotoxin a/b receptor, Wsl-1, TL1A/TNFSF15, BAFF, BAFF-R/TNFRSF13C, TRAIL R2/TNFRSF10B, TRAIL R2/TNFRSF10B, Fas/TNFRSF6 CD27/TNFRSF7, DR3/TNFRSF25, HVEM/TNFRSF14, TROY/TNFRSF19, ligand CD40/TNFSF5, BCMA/TNFRSF17, CD30/TNFRSF8, LIGHT/TNFSF14, 4-1BB/TNFRSF9, CD40/TNFRSF5, GITR/TNFRSF18, osteoprotegerin/TNFRSF11B, RANK/TNFRSF11 A, TRAIL R3/TNFRSF10C, TRAIL/TNFSF10, TRANCE/RANK L/TNFSF11, 4-1BB Ligand/TNFSF9, TWEAK/TNFSF12, CD40/TNFSF5 ligand, Fas/TNFSF6 ligand, RELT/TNFRSF19L, APRIL/TNFSF13, DcR3/TNFRSF6B, TNF RI/TNFRSF1A, TRAIL R1/TNFRSF10A, TRAIL R4/TNFRSF10D, CD30/TNFSF8 ligand, GITR/TNFSF18 ligand , TNFSF18, TACI/TNFRSF13B, NGF R/TNFRSF16, OX40/TNFSF4 ligand, TRAIL R2/TNFRSF10B, TRAIL R3/TNFRSF10C, TWEAK R/TNFRSF12, BAFF/BLyS/TNFSF13, DR6/TNFRSF21, TNF-alpha/TNFSF1A, Pro- TNFalpha/TNFSF1A, lymphotoxin beta R/TNFRSF3, lymphotoxin beta R (LTbR)/Fc chimera, TNF RI/TNFRSF1A, TNF-beta/TNFSF1B, PGRP-S, TNF RI/TNFRSF1A, TNF RII/TNFRSF1B, EDA- A2, TNF-alpha/TNFSF1A, EDAR, XEDAR, TNF RI/TNFRSF1A.

[0211]Of particular interest are human target proteins that are commercially available in purified form. Examples are: 4EBP1, 14-3-3 zeta, 53BP1, 2B4/SLAMF4, CCL21/6Ccina, 4-1BB/TNFRSF9, 8D6A, 4-1BB/TNFSF9 ligand, 8-oxo-dG, 4-amino-1,8 -naphthalimide, A2B5, aminopeptidase LRAP/ERAP2, A33, aminopeptidase N/ANPEP, Aag, aminopeptidase P2/YPNPEP2, ABCG2, aminopeptidase P1/XPNPEP1, ACE, aminopeptidase PILS/ARTS1, ACE-2, Amnionless, actin, amphiregulin, beta- actin, A<m>PK alpha 1/2, activin A, AMPK alpha 1, activin AB, AMPK alpha 2, activin B, AMPK beta 1, activin C, AMPK beta 2, activin RIA/ALK-2, androgen R /NR3C4, activin RIB/ALK-4, angiogenin, activin RIIA, angiopoietin-1, activin RIIB, angiopoietin-2, ADAM8, angiopoietin-3, ADAM9, angiopoietin-4, ADAM10, 1 angiopoietin-like, ADAM12, 2 similar angiopoietin, ADAM15, 3 angiopoietin-like, TACE/ADAM17, 4 angiopoietin-like, ADAM19, 7 angiopoietin-like/CDT6, ADAM33, angiostatin, ADAMTS4, annexin A1/annexin I, AD<a>M<t>S5, annexin A7, ADAMTS1, annexin A10, ADAMTSL-1/punctin, annexin V, adiponectin/Acrp30, ANP, AEBSF, AP site, aggrecan, APAJF-1, agrin, APC, AgRP, APE, A<g>TR-2, APJ , AIF, APLP-1, Akt, APLP-2, Akt1, apolipoprotein AI, Akt2, apolipoprotein B, Akt3, APP, serum albumin, APRIL/TNFSF13, ALCAM, ARC, ALK-1, artemin, ALK-7, arylsulfatase A/ARSA, alkaline phosphatase, ASAH2/N-acylesphingosine amidohydrolase-2, alpha 2u-globulin, ASC, alpha-1-acid glycoprotein, ASGR1, alpha-fetoprotein, ASK1, ALS, ATM, ameloblastin, ATRIP, AMICA/JAML, Aurora A, AMIGO, Aurora B, AMIGO2, axin-1, AMIGO3, Axl, aminoacylase/ACY1, azurocidin/CAP37/HBP, aminopeptidase A/ENPEP, B4 gALT 1, BIM, B7-1/CD80, 6-biotin-17 -NAD, B7-2/CD86, BLAME/SLAMF8, B7-H1/PD-L1, CXCL13/BLC/BCA-1, B7-H2, BLIMP1, B7-H3, Blk, B7-H4, BMI-1, BACE -1, BMP-1/PCP, BACE-2, BMP-2, Bad, BMP-3, BAFF/TNFSF13B, BMP-3b/GDF-10, BAFF R/TNFRSF13C, BMP-4, Bag-1, BMP- 5, BAK, BMP-6, BAMBI/NMA, BMP-7, BARD1, BMP-8, Bax, BMP-9, BCAM, BMP-10, Bcl-10, BMP-15/GDF-9B, Bcl-2, BMPR-IA/ALK-3, Bcl-2-related protein A1, BMPR-IB/ALK-6, Bcl-w, BMPR-II, Bcl-x, BNIP3L, Bcl-xL, BOC, BCMA/TNFRSF17, BOK, BDNF, BPDE, benzamide, Brachyury, common beta chain, B-Raf, beta IG-H3, CXCL14/BRAK, betacellulin, BRCA1, beta-defensin 2, BRCA2, BID, BTLA, biglycan, Bub-1, killer protein-like Bik, c-jun, CD90/Thy1, c-Rel, CD94, CCL6/C10, CD97, Clq R1/CD93, CD151, ClqTNF1, CD160, ClqTNF4, CD163, ClqTNF5, CD164, complement component C1r, CD200, complement component complement C1s, CD200 R1, complement component C2,<c>D229/SLAMF3, complement component C3a, CD23/Fc epsilon RII, complement component C3d, CD2F-10/SLAMF9, complement component C5a, CD5L, cadherin-4 /R-cadherin, CD69, cadherin-6, CDC2, cadherin-8, CDC25A, cadherin-11, CDC25B, cadherin-12, CDCP1, cadherin-13, CDO, cadherin-17, CDX4, E-cadherin, CEACAM-1 /CD66a, N-cadherin, CEACAM-6, P-cadherin, Cerberus 1, VE-cadherin, CFTR, calbindin D, cGMP, calcineurin A, Chem R23, calcineurin B, Chemerin, calreticulin-2, chemokine sampling containers, CaM kinase II, chitinase 3-like 1, cAMP, chitotriosidase/CHITI, cannabinoid R1; Chk1, cannabinoid R2/CB2/CNR2, Chk2, CAR/NR1I3, CHL-1/L1CAM-2, carbonic anhydrase I, choline acetyltransferase/ChAT, carbonic anhydrase II, chondrolectin, carbonic anhydrase III, chordin, carbonic anhydrase IV, 1 chordin-like, carbonic anhydrase VA, chordin-like 2, carbonic anhydrase VB, CINC-1, carbonic anhydrase VI, CINC-2, carbonic anhydrase VII, CINC-3, carbonic anhydrase VIII, claspin, carbonic anhydrase IX, claudin-6 , carbonic anhydrase X, CLC, carbonic anhydrase XII, CLEC-1, carbonic anhydrase XIII, CLEC-2, carbonic anhydrase P1/COLEC12, carboxypeptidase A4, clusterin, carboxypeptidase B1, clusterin-like 1, carboxypeptidase E/CPE, CMG-2, carboxypeptidase , CNT<f>R alpha, caspase, coagulation factor II/thrombin, caspase-1, coagulation factor III/tissue factor, caspase-2, coagulation factor VII, caspase-3, coagulation factor X, caspase-4 , coagulation factor XI, caspase-6, coagulation factor , caspase-13, common gamma chain/IL-2 R gamma, caspase peptide inhibitors, COMP/thrombospondin-5, catalase, complement component C1rLP, beta-catenin, complement component ClqA, cathepsin 1, complement component C1qC , cathepsin 3, complement factor D, cathepsin 6, complement factor I, cathepsin A, complement MASP3, cathepsin B, connexin 43, cathepsin C/DPPI, contactin-1, cathepsin D, contactin-2/TAG1, cathepsin E, contactin-4, cathepsin F, contactin-5, cathepsin H, corin, cathepsin L, comulin, cathepsin O, CORS26/ClqTNF,3, cathepsin S, rat cortical stem cells, cathepsin V, cortisol, cathepsin X/Z/P, COUP-TF I/NR2F1, CBP, COUP-TF II/NR2F2, CCI, COX-1, CCK-A R, COX-2, CCL28, CRACC/SLAMF7, CCR1, C-reactive protein, CCR2, creatine kinase, muscle/ CKMM, CCR3, creatinine, CCR4, CREB, CCR5, CREG, CCR6, CRELD1, CCR7, CRELD2, CCR8, CRHBP, CCR9, CRHR-1, CCR10, CRIM1, CD155/PVR, Cripto, CD2, CRISP-2, CD3, CRISP-3, CD4, Crossveinless-2, CD4+/45RA-, CRTAM, CD4+/45RO-, CRTH-2, CD4+/CD62L-/CD44, CRY1, CD4+/CD62L+/CD44, Cryptic, CD5, CSB/ERCC6, CD6 , CCL27/CTACK, CD8, CTGF/CCN2, CD8+/45RA-, CTLA-4, CD8+/45RO-, cubilin, CD9, CX3CR1, CD14, CXADP, CD27/TNFRSF7, CXCL16, CD27/TNFSF7 ligand, CXCR3, CD28, CXCR4, CD30/TNFRSF8, CXCR5, CD30/TNFSF8 ligand, CXCR6, CD31/PECAM-1, cyclophilin A, CD34, Cyr61/CCN1, CD36/SR-B3, cystatin A, CD38, cystatin B, CD40/TNFRSF5, cystatin C , CD40/TNFSF5 ligand, cystatin D, CD43, cystatin E/M, CD44, cystatin F, CD45, cystatin H, CD46, cystatin H2, CD47, cystatin S, CD48/SLAMF2, cystatin SA, CD55/DAF, cystatin SN, CD58/LFA-3, cytochrome c, CD59, apocytochrome c, CD68, holocytochrome c, CD72, cytokeratin 8, CD74, cytokeratin 14, CD83, cytokeratin 19, CD84/SLAMF5, cytonin, D6, DISP1, DAN, Dkk-1, DANCE, Dkk-2, DARPP-32, Dkk-3, DAX1/NR0B1, Dkk-4, DCC, DLEC, DCIR/CLEC4A, DLL1, DCAR, DLL4, DcR3/TNFRSF6B, d-luciferin, DC-SIGN, DNA ligase IV, DC-SIGNR/CD299, DNA polymerase beta, DcTRAIL R1/TNFRSF23, ADNM-1, DcTRAIL R2/TNFRSF22, DNA-PKcs, DDR1, DNER, DDR2, dopa decarboxylase/DDC, DEC-205, DPCR-1, decapentaplegic , DPP6, decorin, DPPA4, dectin-1/CLEC7A, DPPA5/ESG1, dectin-2/CLEC6A, DPPII/QPP/DPP7, DEP-1/CD148, DPPIV/CD26, Desert Hedgehog, DR3/TNFRSF25, desmin, DR6/ TNFRSF21, desmoglein-1, DSCAM, desmoglein-2, DSCAM-L1, desmoglein-3, DSPG3, Disheveled-1, Dtk, Disheveled-3, dynamin, EAR2/NR2F6, EphA5, ECE-1, EphA6, ECE-2, EphA7, ECF-L/CHL3L3, EphA8, ECM-1, EphB1, ecotin, EphB2, EDA, EphB3, EDA-A2, EphB4, EDAR, EphB6, EDG-1, ephrin, EDG-5, ephrin-A1, EDG- 8, ephrin-A2, eEF-2, ephrin-A3, EGF, ephrin-A4, EGF R, ephrin-A5, EGR1, ephrin-B, EG-VEGF/PK1, ephrin-B1, eIF2 alpha, ephrin-B2, eIF4E, ephrin-B3, Elk-1, epigene, EMAP-II, epimorphin/syntaxin 2, EMMPRIN/CD147, epiregulin, CXCL5/ENA, EPR-1/Xa receptor, endocan, ErbB2, endoglin/CD105, ErbB3, endoglycan, ErbB4, endonuclease III, ERCC1, endonuclease IV, ERCC3, endonuclease V, ERK1/ERK2, endonuclease VIII, ERK1, endorepelin/perlecan, ERK2, endostatin, ERK3, endothelin-1, ERK5/BMK1, Engrailed-2, ERR alpha/NR3B1 , EN-RAGE, ERR beta/NR3B2, enteropeptidase/enterokinase, ERR gamma/NR3B3, CCL11/eotaxin, erythropoietin, CCL24/eotaxin-2, erythropoietin R, CCL26/eotaxin-3, ESAM, EpCAM/TROP-1, ER alpha /NR3A1, EPCR, ER beta/NR3A2, Eph, exonuclease III, EphA1, exostosin-like 2/EXTL2, EphA2, exostosin-like 3 /EXTL3, EphA3, FABP1, FGF-BP, FABP2, FGF R1-4, FABP3, FGF R1, FABP4, FGF R2, FABP5, FGF R3, FABP7, FGF R4, FABP9, FGF R5, complement factor B, Fgr, FADD, FHR5, FAM3A, fibronectin, FAM3B, ficolin-2, FAM3C, ficolin-3, FAM3D, FITC, fibroblast activation protein alpha/FAP, FKBP38, Fas/TNFRSF6, Flap, Fas ligand /TNFSF6, FLIP, FATP1, FLRG, FATP4, FLRT1, FATP5, FLRT2, Fc gamma RI/CD64, FLRT3, Fc gamma RIIB/CD32b, Flt-3, Fc gamma RIIC/CD32c, Flt-3 ligand, Fc gamma RIIA/CD32a, follistatin, Fc gamma RIII/CD16, follistatin-like 1, FcRH1/IRTA5, FosB/G0S3, FcRH2/IRTA4, FoxD3, FcRH4/IRTA1, FoxJ1, FcRH5/IRTA2, FoxP3, Fc receptor-like 3/CD16-2, Fpg, FEN-1, FPR1, fetuin A, FPRL1, fetuin B, FPRL2, acidic FGF, CX3CL1/fractalkine, FGF basic, Frizzled-1, FGF-3, Frizzled-2, FGF-4, Frizzled-3, FGF-5. Frizzled-4, FGF-6, Frizzled-5, FGF-8, Frizzled-6, FGF-9, Frizzled-7, FGF-10, Frizzled-8, FGF-11, Frizzled-9, FGF-12, Frk, FGF-13, sFRP-1, FGF-16, sFRP-2, FGF-17, sFRP-3, FGF-19, sFRP-4, FGF-20, furin, FGF-21, FXR/NR1H4, FGF-22, Fyn, FGF-23, G9a/EHMT2, GFR alpha-3/GDNF R alpha-3, GABA-A-R alpha 1, GFR alpha-4/GDNF R alpha-4, GABA-A-R alpha 2, GITR/TNFRSF18, GABA- A-R alpha 4, GITR/TNFSF18 ligand, GABA-A-R alpha 5, GLI-1, GABA-A-R alpha 6, GLI-2, GABA-A-R beta 1, GLP/EHMT1, GABA-A-R beta 2, GLP-1 R, GABA-A-R beta 3, glucagon, GABA-A-R gamma 2, glucosamine (N-acetyl)-6-sulfatase/GNS, GABA-B-R2, GluR1, GAD1/GAD67, GluR2/3, GAD2/GAD65, GluR2, GADD45 alpha, GluR3, GADD45 beta, Glut1, GADD45 gamma, Glut2, galectin-1, Glut3, galectin-2, Glut4, galectin-3, Glut5, galectin-3 BP, glutaredoxin 1, galectin-4, glycine R, galectin-7 , glycophorin A, galectin-8, glypican 2, galectin-9, glypican 3, GaINAc4S-6ST, glypican 5, GAP-43, glypican 6, GAPDH, GM-CSF, Gas1, GM-CSF R alpha, Gas6, GMF- beta, GASP-1/WFIKKNRP, gp130, GASP-2/WFIKKN, glycogen phosphorylase BB/GPBB, GATA-1, GPR15, GATA-2, GPR39, GATA-3, GPVI, GATA-4, GR/NR3C1, GATA- 5, Gr-1/Ly-6G, GATA-6, granulysin, GBL, granzyme A, GCNT/NR6A 1, granzyme B, CXCL6/GCP-2, granzyme D, GCSF, granzyme G, G-CSF R, granzyme H , GDF-1, GRASP, GDF-3 GRB2, GDF-5, Gremlin, GDF-6, GRO, GDF-7, CXCL1/GRO alpha, GDF-8, CXCL2/GRO beta, GDF-9, CXCL3/GRO gamma , GDF-11, growth hormone, GDF-15, growth hormone R, GDNF, GRP75/HSPA9B, GFAP, GSK-3 alpha/beta, GFI-1, GSK-3 alpha, GFR alpha-1/GDNF R alpha -1, GSK-3 beta, GFR alpha-2/GDNF R alpha-2, EZFIT, H2AX, histidine, H60, HM74A, HAI-1, HMGA2, HAI-2, HMGB1, HAI-2A, TCF-2/HNF -1 beta, HAI-2B, HNF-3 beta/FoxA2, HAND1, HNF-4 alpha/NR2A1, HAPLN1, HNF-4 gamma/NR2A2, airway trypsin-like protease/HAT, HO-1/HMOX1/ HSP32, HB-EGF, HO-2/HMOX2, CCL14a/HCC-1, HPRG, CCL14b/HCC-3, Hrk, CCL16/HCC-4, HRP-1, alpha HCG, HS6ST2, Hck HSD-1, HCR/ CRAM-A/B, HSD-2, HDGF, HSP10/EPF, Hemoglobin, HSP27, Hepassocin, HSP60, HES-1, HSP70, HES-4, HSP90, HGF, HTRA/protease Do, HGF activator, HTRA1/PRSS11 , HGF R, HTRA2/Omi, HIF-1 alpha, HVEM/TNFRSF14, HIF-2 alpha, hyaluronan, HIN-1/secretoglobulin 3A1, 4-hydroxynonenal, Hip,CCL1/I-309/TCA-3, IL-10 , cIAP (pan), IL-10 R alpha, cIAP-1/HIAP-2, IL-10 R beta, cIAP-2/HLAP-1, IL-11, IBSP/sialoprotein II, IL-11 R alpha, ICAM -1/CD54, IL-12, ICAM-2/CD102, IL-12/IL-23 p40, ICAM-3/CD50, IL-12 R beta 1, ICAM-5, IL-12 R beta 2, ICAT, IL-13, ICOS, IL-13 R alpha 1, iduronate 2-sulfatase/IDS, IL-13 R alpha 2, IFN, IL-15, IFN-alpha, IL-15 R alpha, IFN-alpha 1, IL- 16, IFN-alpha 2, IL-17, IFN-alpha 4b, IL-17 R, IFN-alpha A, IL-17 RC, IFN-alpha B2, IL-17 RD, IFN-alpha C, 1L-17B, IFN-alpha D, IL-17B R, IFN-alpha F, IL-17C, IFN-alpha G, IL-17D, IFN-alpha H2, IL-17E, IFN-alpha I, IL-17F, IFN-alpha J1 , IL-18/IL-IF4, IFN-alpha K, IL-18 BPa, IFN-alpha WA, IL-18 BPc, IFN-alpha/beta R1, IL-18 BPd, IFN-alpha/beta R2, IL- 18 R alpha/IL-1 R5, IFN-beta, IL-18 R beta/IL-1 R7, IFN-gamma, IL-19, IFN-gamma R1, IL-20, IFN-gamma R2, IL-20 R alpha, IFN-omega, IL-20 R beta, IgE, IL-21, IGFBP-1, IL-21 R, IGFBP-2, IL-22, IGFBP-3, IL-22 R, IGFBP-4, IL- 22BP, IGFBP-5, IL-23, IGFBP-6, IL-23 R, IGFBP-L1, IL-24, IGFBP-rp1/IGFBP-7, IL-26/AK155, IGFBP-rP10, IL-27, IGF -I, IL-28A, IGF-I R, IL-28B, IGF-II, IL-29/IFN-lambda 1, IGF-II R, IL-31, IgG, IL-31 RA, IgM, IL-32 alpha, IGSF2, IL-33, IGSF4A/SynCAM, ILT2/CD85j, IGSF4B, ILT3/CD85k, IGSF8, ILT4/CD85d, IgY, ILT5/CD85a, IkB-beta, ILT6/CD85e, IKK alfa, Indian Hedgehog, IKK epsilon , INSRR, IKK gamma, insulin, IL-1 alpha/IL-1F1, insulin R/CD220, IL-1 beta/IL-1F2, proinsulin, IL-1 ra/IL-1 F3, insulisin/IDE, 1L-1F5 /FIL1 delta, integrin alpha 2/CD49b, IL-1F6/FIL1 epsilon, integrin alpha 3/CD49c, IL-1F7/FIL1 zeta, integrin alpha 3 beta 1/VLA-3, IL-1F8/FIL1 eta, integrin alpha 4 /CD49d, IL-1F9/IL-1 H1, integrin alpha 5/CD49e, IL-1F10/IL-1HY2, integrin alpha 5 beta 1, IL-1 RI, integrin alpha 6/CD49f, IL-1 RII, integrin alpha 7, IL-1 R3/IL-1 R AcP, integrin alpha 9, IL-1 R4/ST2, integrin alpha E/CD103, IL-1 R6/IL-1 R rp2; integrin alpha L/CD11a, IL-1 R8, integrin alpha L beta 2, IL-1 R9, integrin alpha M/CD11b, IL-2, integrin alpha M beta 2, IL-2 R alpha, integrin alpha V/CD51, IL-2 R beta, integrin alpha V beta 5, IL-3, integrin alpha V beta 3, IL-3 R alpha, integrin alpha V beta 6, IL-3 R beta, integrin alpha X/CD11c, IL-4, integrin beta 1/CD29, IL-4 R, integrin beta 2/CD18, IL-5, integrin beta 3/CD61, IL-5 R alpha, integrin beta 5, IL-6, integrin beta 6, IL-6 R, integrin beta 7, IL-7, CXCL10/IP-10/CRG-2, IL-7 R alpha/CD127, IRAK1, CXCR1/IL-8 RA, IRAK4, CXCR2/IL-8 RB, IRS-1, CXCL8/ IL-8, Islet-1, IL-9, CXCL11/1-TAC, IL-9 R, Jagged 1, JAM-4/IGSF5, Jagged 2, JNK, JAM-A, JNK1/JNK2, JAM-B/VE -JAM, JNK1, JAM-C, JNK2, kininogen, kallikrein 3/PSA, quininostatin, kallikrein 4, KIR/CD158, kallikrein 5, KIR2DL1, kallikrein 6/neurosin, KIR2DL3, kallikrein 7, KIR2DLA/CD158d, kallikrein 8/neuropsin , KIR2DS4, kallikrein 9, KIR3DL1, plasma kallikrein/KLKB1, KIR3DL2, kallikrein 10, Kirrel2, kallikrein 11, KLF4, kallikrein 12, KLF5, kallikrein 13, KLF6, kallikrein 14, Klotho, kallikrein 15, Klotho beta, KC, KOR, Keap1, Kremen-1, Kell, Kremen-2, KGF/FGF-7, LAG-3, LINGO-2, LAIR1, lipin 2, LAIR2, lipocalin-1, laminin alpha 4, lipocalin-2/NGAL, laminin gamma 1 , 5-lipoxygenase, laminin I, LXR alpha/NR1H3, laminin S, LXR beta/NR1H2, laminin-1, livin, laminin-5, LIX, LAMP, LMIR1/CD300A, langerin, LMIR2/CD300c, LAR, LMIR3/CD300LF , latexin, LMIRS/CD300LB, layilin, LMIR6/CD300LE, LBP, LM02, LDL R, LOX-1/SR-E1, LECT2, LRH-1/NR5A2, LEDGF, LRIG1, Lefty, LRIG3, Lefty-1, LRP- 1, Lefty-A, LRP-6, legumain, LSECtin/CLEC4G, leptin, lumican, leptin R, CXCL15/Lungkine, leukotriene B4, , lymphotoxin beta R/TNFRSF3, LIGHT/TNFSF14, Lyn, limitin, Lyp, LIMPII/SR-B2, lysyl oxidase homolog 2, LIN-28, LYVE-1, LINGO-1, alpha 2-macroglobulin, CXCL9/MIG, MAD2L1, mimecane, MAdCAM-1, mindin, MafB, mineralocorticoid R/NR3C2, MafF, CCL3L1/MIP-1 alpha isoform LD78 beta, MafG, CCL3/MIP-1 alpha, MafK, CCL4L1/LAG-1, MAG/Siglec -4a, CCL4/MIP-1 beta, MANF, CCL15/MIP-1 delta, MAP2, CCL9/10/MIP-1 gamma, MAPK, MIP-2, marapsin/pancreasin, CCL19/MIP-3 beta, MARCKS, CCL20 /MIP-3 alpha, MARCO, MIP-I, Mash1, MIP-II, Matrilin-2, MIP-III, Matrilin-3, MIS/AMH, Matrilin-4, MIS RII, Matriptase/ST14, MIXL1, MBL, MKK3 /MKK6, MBL-2, MKK3, melanocortin 3R/MC3R, MKK4, MCAM/CD146, MKK6, MCK-2, MKK7, Mcl-1, MKP-3, MCP-6, MLH-1, CCL2/MCP-1, MLK4 alpha, MCP-11, MMP, CCL8/MCP-2, MMP-1, CCL7/MCP-3/MARC, MMP-2, CCL13/MCP-4, MMP-3, CCL12/MCP-5, MMP-7 , M-CSF, MMP-8, M-CSF R, MMP-9, MCV-type II, MMP-10, MD-1, MMP-11, MD-2, MMP-12, CCL22/MDC, MMP-13 , MDL-1/CLEC5A, MMP-14, MDM2, MMP-15, MEA-1, MMP-16/MT3-MMP, MEK1/MEK2, MMP-24/MT5-MMP, MEK1, MMP-25/MT6-MMP , MEK2, MMP-26, melusin, MMR, MEPE, MOG, meprin alpha, CCL23/MPBP-1, meprin beta, M-Ras/R-Ras3, Mer, Mre11, mesothelin, MRP1, meteorin, MSK1/MSK2, methionine aminopeptidase 1, MSK1, methionine aminopeptidase, MSK2, methionine aminopeptidase 2, MSP, MFG-E8, MSP R/Ron, MFRP, Mug, MgcRacGAP, MULT-1, MGL2, Musashi-1, MGMT, Musashi-2, MIA, MuSK , MICA, MutY DNA glycosylase, MICB, MyD88, MICL/CLEC12A, myeloperoxidase, beta 2 microglobulin, myocardin, Midkine, myocilin, MIF, myoglobin, NAIP NGFI-B gamma/NR4A3, Nanog, NgR2/NgRH1, CXCL7/NAP-2 , NgR3/NgRH2, Nbsl, nidogen-1/entactin, NCAM-1/CD56, nidogen-2, NCAM-L1, nitric oxide, nectin-1, nitrotyrosine, nectin-2/CD112, NKG2A, nectin-3, NKG2C, nectin-4, NKG2D, neogenin, NKp30, neprilysin/CD10, NKp44, neprilysin-2/MMEL1/MMEL2, NKp46/NCR1, nestin, NKp80/KLRF1, NET02, NKX2.5, netrin-1, NMDA R, NR1 subunit , netrin-2, N<m>D<a>R, NR2A subunit, netrin-4, NMDA R, NR2B subunit, netrin-G1a, NMDA R, NR2C subunit, netrin-G2a, N-Me-6 ,7-diOH-TIQ, neuregulin-1/NRG1, nodal, neuregulin-3/NRG3, noggin, neuritin, nogo receptor, NeuroD1, Nogo-A, neurofascin, NOMO, neurogenin-1, nope, neurogenin-2, norrin , neurogenin-3, eNOS, neurolysin, iNOS, neurophysin II, nNOS, neurophilin-1, Notch-1, neurophilin-2, Notch-2, neuropoietin, Notch-3, neurotrimin, Notch-4, neurturin, NOV/CCN3, NFAM1, NRAGE, NF-H, NrCAM, NFkB1, NRL, NFkB2, NT-3, NF-L, NT-4, NF-M, NTB-A/SLAMF6, NG2/MCSP, NTH1, NGF R/TNFRSF16, nucleostemin , beta-NGF, Nurr-1/NR4A2, NGFI-B alpha/NR4A1, OAS2, orexin B, OBCAM, OSCAR, OCAM, OSF-2/periostin, OCIL/CLEC2d, oncostatin M/OSM, OCILRP2/CLEC2i, OSM R beta, Oct-3/4, osteoactivin/GPNMB, OGG1, osteoadherin, Olig 1, 2, 3, osteocalcin, Olig 1, osteocrin, Olig2, osteopontin, Olig3, osteoprotegerin/TNFRSF11B, oligodendrocyte marker O1, Otx2, oligodendrocyte marker 04, OV-6, OMgp, OX40/TNFRSF4, opticin, OX40/TNFSF4 ligand, orexin A, OAS2, orexin B, OBCAM, OSCAR, Ocam, OSF-2/periostin, OCIL/CLEC2d, oncostatin M/OSM, OCILRP2/ CLEC2i, OSM R beta, Oct-3/4, osteoactivin/GPNMB, OGG1, osteoadherin, Olig 1, 2, 3, osteocalcin, Olig1, osteocrin, Olig2, osteopontin, Olig3, osteoprotegerin/TNFRSF11B, oligodendrocyte marker O1, Otx2, oligodendrocyte marker 04, OV-6, OMgp, OX40/TNFRSF4, opticin, OX40/TNFSF4 ligand, orexin A, RACK1, Ret, Rad1, REV-ERB alpha/NR1D1, Rad17, REV-ERB beta/NR1D2, Rad51, Rex -1, Rae-1, RGM-A, Rae-1 alpha, RGM-B, Rae-1 beta, RGM-C, Rae-1 delta, Rheb, Rae-1 epsilon, ribosomal protein S6, Rae-1 gamma, RIP1, Raf-1, ROBO1, RAGE, ROB02, Ra1A/Ra1B, ROB03, Ra1A, ROBO4, Ra1B, ROR/NR1F1-3 (pan), RANK/TNFRSF11A, ROR alpha/NR1F1, CCL5/RANTES, ROR gamma/NR1F3 , Rap1A/B, RTK/ROR1-like orphan receptor 1, RAR alpha/NR1B1, RTK/ROR2-like orphan receptor 2, R<a>R beta/NR1B2, RP105, RAR gamma/NR1B3, RPA2, Ras, RSK (pan), RBP4, RSK1/RSK2, RECK, RSK1, Reg 2/PAP, RSK2, Reg I, RSK3, Reg II, RSK4, Reg III, R-spondin 1, Reg IIIa, R-spondin 2, Reg IV , R-spondin 3, relaxin-1, RUNX1/CBFA2, relaxin-2, RUNX2/CBFA1, relaxin-3, RUNX3/CBFA3, RELM alpha, RXR alpha/NR2B1, RELM beta, RXR beta/NR2B2, RELT/TNFRSF19L, RXR gamma/NR2B3, resistin, S100A10, SLITRK5, S100A8, SLPI, S100A9, SMAC/diablo, S100B, Smad1, S100P, Smad2, SALL1, Smad3, delta-sarcoglycan, Smad4, Sca-1/Ly6, Smad5, SCD-1 , Smad7, SCF, Smad8, SCF R/c-kit, SMC1, SCGF, alpha-smooth muscle actin, SCL/Tal1, SMUG1, SCP3/SYCP3, Snail, CXCL12/SDF-1, sodium calcium exchanger 1, SDNSF/MCFD2, Soggy-1, alpha-secretase, Sonic Hedgehog, gamma-secretase, SorCS1, betasecretase, SorCS3, E-selectin, sortilin, L-selectin, SOST, P-selectin, SOX1, semaphorin 3A, SOX2, semaphorin 3C , SOX3, semaphorin 3E, SOX7, semaphorin 3F, SOX9, semaphorin 6A, SOX10, semaphorin 6B, SOX17, semaphorin 6C, SOX21 semaphorin 6D,SPARC, semaphorin 7A, SPARC-like 1, separase, SP-D, serine substrate I /threonine phosphatase, spinesin, serpin A1, F-spondin, serpin A3, SR-AI/MSR, serpin A4/kallistatin, Src, serpin A5/protein C inhibitor, SREC-I/SR-F1, serpin A8/angiotensinogen, SREC-II, serpin B5, SSEA-1, serpin C1/antithrombin-III, SSEA-3, serpin D1/heparin cofactor II, SSEA-4, serpin E1/PAI-1, ST7/LRP12, serpin E2, stabilin- 1, Serpin F1, Stabilin-2, Serpin F2, Staniocalcin 1, Serpin G1/C1 inhibitor, Staniocalcin 2, Serpin I2, STAT1, Serum amyloid A1, STAT2, SF-1/NR5A1, STAT3, SGK, STAT4, SHBG , STAT5a/b, SHIP, STAT5a, SHP/NR0B2, STAT5b, SHP-1, STAT6, SHP-2, VE-statin, SIGIRR, Stella/Dppa3, Siglec-2/CD22, STRO-1, Siglec-3/CD33 , Substance P, Siglec-5, Sulfamidase/SGSH, Siglec-6, Sulfatase Modifying Factor 1/SUMF1, Siglec-7, Sulfatase Modifying Factor 2/SUMF2, Siglec-9, SUMO1, Siglec-10, SUMO2/3/ 4, Siglec-11, SUMO3, Siglec-F, superoxide dismutase, SIGNR1/CD209, superoxide dismutase-1/Cu-Zn SOD, SIGNR4, superoxide dismutase-2/Mn-SOD, SIRP beta 1, superoxide dismutase-3/EC -SOD, SKI, survivin, SLAM/CD150, synapsin I, Sleeping Beauty transposase, syndecan-1/CD138, Slit3, syndecan-2, SLITRK1, syndecan-3, SLITRK2, syndecan-4, SLITRK4, TACI/TNFRSF13B, TMEFF1/ tomoregulin-1, TAO2, TMEFF2, TAPP1, TNF-alpha/TNFSF1A, CCL17/TARC, TNF-beta/TNFSF1B, Tau, TNF RI/TNFRSF1A.TC21/R-Ras2, TNF RII/TNFRSF1B, TCAM-1, TOR, TCCR/WSX-1, TP-1, TC-PTP, TP63/TP73L, TDG, TR, CCL25/TECK, TR alpha/NR1A1, tenascin C, TR beta 1/NR1A2, tenascin R, TR2/NR2C1, TER-119 , TR4/NR2C2, TERT, TRA-1-85, testican 1/SPOCK1, TRADD, testican 2/SPOCK2,TRAF-1, testican 3/SPOCK3, TRAF-2, TFPI, TRAF-3, TFPI-2, TRAF- 4, TGF-alpha, TRAF-6, TGF-beta, TRAIL/TNFSF10, TGF-beta 1, TRAIL R1/TNFRSF10A, LAP (TGF-beta 1), TRAIL R2/TNFRSF10B, latent TGF-beta 1, TRAIL R3/ TNFRSF10C, TGF-beta 1,2, TRAIL R4/TNFRSF10D, TGF-beta 2, TRANCE/TNFSF11, TGF-beta 3, TfR (transferrin R), TGF-beta 5, apo-transferrin, latent TGF-beta pb 1, holo-transferrin, latent TGF-beta pb 2, trapin-2/elafin, latent TGF-beta pb4, TREM-1, TGF-beta RI/ALK-5, TREM-2, TGF-beta RII, TREM-3, TGF -beta RIIb, TREML1/TLT-1, TGF-beta RIII, TRF-1, thermolysin, TRF-2, thioredoxin-1, TRH/TRHDE-degrading ectoenzyme, thioredoxin-2, TRIM5, thioredoxin-80, tripeptidyl peptidase I , thioredoxin-like /TRP14, TrkA, THOP1, TrkB, thrombomodulin/CD141, TrkC, thrombopoietin, TROP-2, thrombopoietin R, troponin I peptide 3, thrombospondin-1, troponin T, thrombospondin-2, TROY/TNFRSF19, thrombospondin-4, trypsin 1, thymopoietin, trypsin 2/PRSS2, thymic chemokine-1, trypsin 3/PRSS3, Tie-1, tryptase-5/Prss32, Tie-2, tryptase alpha/TPS1, TIM-1/KIM- 1/HAVCR, tryptase beta-1/MCPT-7, TIM-2, tryptase beta-2/TPSB2, TIM-3, tryptase epsilon/BSSP-4, TIM-4, tryptase gamma-1/TPSG1, TIM-5, tryptophan hydroxylase, TIM-6, TSC22, TIMP-1, TSG, TTMP-2, TSG-6, TIMP-3, TSK, TIMP-4, TSLP, TL1A/TNFSF15, TSLP R, TLR1, TSP50, TLR2, beta- III tubulin, TLR3, TWEAK/TNFSF12, TLR4, TWEAK R/TNFRSF12, TLR5, Tyk2, TLR6, phospho-tyrosine, TLR9, tyrosine hydroxytase, TLX/NR2E1, tyrosine phosphatase substrate I, ubiquitin, UNCSH3, Ugi, UNCSH4, UGRP1 , UNG, ULBP-1, uPA, ULBP-2, uPAR, ULBP-3, URB, UNC5H1, UVDE, UNC5H2, vanilloid R1, VEGF R, VASA, VEGF R1/Flt-1, vasohibin, VEGF R2/KDR/Flk -1, vasorin, VEGF R3/Flt-4, vasostatin, versican, Vav-1, VG5Q, VCAM-1, VHR, VDR/NR1I1, vimentin, VEGF, vitronectin, VEGF-B, VLDLR, VEGF-C, vWF- A2, VEGF-D, alpha-synuclein, Ku70, WASP, Wnt-7b, WIF-1, Wnt-8a WISP-1/CCN4, Wnt-8b, WNK1, Wnt-9a, Wnt-1, Wnt-9b, Wnt -3a, Wnt-10a, Wnt-4, Wnt-10b, Wnt-5a, Wnt-11, Wnt-5b,wnvNS3, Wnt7a, XCR1, XPE/DDB1, XEDAR, XPE/DDB2, Xg, XPF, XIAP, XPG , XPA, XPV, XPD, XRCC1, Yes, YY1, EphA4.

[0212]Numerous human ion channels are targets of particular interest. Non-limiting examples include 5-hydroxytryptamine receptor 3 subunit B, 5-hydroxytryptamine receptor 3 precursor, 5-hydroxytryptamine receptor 3 subunit C, AAD protein 14, acetylcholine receptor protein, alpha subunit precursor, acetylcholine receptor protein, beta subunit precursor, acetylcholine receptor protein, delta subunit precursor, acetylcholine receptor protein, epsilon subunit precursor, acetylcholine receptor protein, gamma subunit precursor, variant acid-sensitive ion channel b-splicing variant 3, acid-sensitive ion channel c-splicing variant 3, acid-sensitive ion channel 4, ADP-ribose pyrophosphatase, mitochondrial precursor, voltage-gated calcium channel alpha1A, amiloride-responsive cation channel 1, neuronal, amiloride-responsive cation channel 2, neuronal amiloride-responsive cation channel 4, isoform 2, amiloride-responsive sodium channel, amiloride-responsive sodium channel alpha subunit, subunit amiloride-sensitive sodium channel beta, amiloride-sensitive sodium channel delta subunit, amiloride-sensitive sodium channel gamma subunit, annexin A7, apical-like protein, ATP-sensitive inwardly rectifying potassium channel 1, ATP-sensitive inwardly rectifying potassium channel 10, ATP-sensitive inwardly rectifying potassium channel 11, ATP-sensitive inwardly rectifying potassium channel 14, ATP-sensitive inwardly rectifying potassium channel 15, ATP-sensitive inwardly rectifying potassium channel ATP 8, alpha 12.2 calcium channel subunit, alpha 12.2 calcium channel subunit, alpha1E calcium channel subunit, splice variant delta19 delta40 delta46, alpha calcium activated potassium channel subunit 1, alpha channel subunit 1 beta calcium activated potassium channels, beta calcium activated potassium channel subunit 2, beta calcium activated potassium channel subunit 3, calcium dependent chloride channel 1, TRPM4B cation channel, CADN fLj90453 fis, clone NT2RP3001542, highly potassium channel tetramerization domain-containing similar 6, CADN FLJ90663 fis, clone PLACE1005031, highly similar to intracellular chloride channel protein 5, beta subunit of CGMP-regulated cation channels, chloride channel protein, channel protein chloride channel protein 2, chloride channel protein 3, chloride channel protein 4, chloride channel protein 5, chloride channel protein 6, chloride channel protein C1C-Ka, chloride channel protein C1C-Kb, chloride channel protein, skeletal muscle, intracellular chloride channel 6, intracellular chloride channel protein 3, intracellular chloride channel protein 4, intracellular chloride channel protein 5, CHRNA3 protein, Clcn3e protein, CLCNKB protein, CNGA4 protein , cullin-5, cyclic g Mp-regulated potassium channel, cyclic nucleotide-regulated 4 cation channel, alpha 3-cyclic nucleotide-regulated cation channel, beta 3-cyclic nucleotide-regulated cation channel, cyclic nucleotide-regulated olfactory channel, regulator of cystic fibrosis transmembrane conductance, cytochrome B-245 heavy chain, dihydropyridine-sensitive L-type, precursor of alpha-2/delta subunits of calcium channels, precursor of FXYD domain-containing ion transport regulator 3, FXYD domain-containing ion transport regulator 5 precursor, FXYD domain-containing ion transport regulator 6 precursor, FXYD domain-containing ion transport regulator 7 precursor, FXYD domain-containing ion transport regulator 8 precursor FXYD domain, G protein-activated inwardly rectifying potassium channel 1, G protein-activated inwardly rectifying potassium channel 2, G protein-activated inwardly rectifying potassium channel 3, G protein-activated inwardly rectifying potassium channel 4, precursor of gamma-aminobutyric acid receptor alpha-1 subunit, precursor of gamma-aminobutyric acid receptor alpha-2 subunit, precursor of gamma-aminobutyric acid receptor alpha-3 subunit, precursor of gamma-aminobutyric acid receptor alpha-4 subunit, precursor of the gamma-aminobutyric acid receptor alpha-5 subunit, precursor of the gamma-aminobutyric acid receptor alpha-6 subunit, precursor of the gamma-aminobutyric acid receptor beta-1 subunit gamma-aminobutyric acid, precursor of the beta-2 subunit of the gamma-aminobutyric acid receptor, precursor of the beta-3 subunit of the gamma-aminobutyric acid receptor, precursor of the delta subunit of the gamma-aminobutyric acid receptor, precursor of the gamma-aminobutyric acid receptor epsilon subunit, precursor of the gamma-aminobutyric acid receptor gamma-1 subunit, precursor of the gamma-aminobutyric acid receptor gamma-3 subunit, precursor of the acid receptor pi subunit gamma-aminobutyric acid, precursor of the rho-1 subunit of the gamma-aminobutyric acid receptor, precursor of the rho-2 subunit of the gamma-aminobutyric acid receptor, precursor of the theta subunit of the gamma-aminobutyric acid receptor, kainate receptor GluR6, glutamate receptor 1 precursor, glutamate receptor 2 precursor, glutamate receptor 3 precursor, glutamate receptor 4 precursor, glutamate receptor 7, glutamate receptor B, glutamate receptor delta-1 subunit precursor glutamate, glutamate receptor, ionotropic kainate 1 precursor, glutamate receptor, ionotropic kainate 2 precursor, glutamate receptor, ionotropic kainate 3 precursor, glutamate receptor, ionotropic kainate 4 precursor, glutamate receptor, ionotropic kainate 5 precursor ionotropic, precursor of glutamate receptor subunit 3A [NMDA], precursor of glutamate receptor subunit 3B [NMDA], precursor of glutamate receptor epsilon 1 subunit [NMDA], precursor of receptor epsilon 2 subunit glutamate [NMDA], precursor of the epsilon 4 subunit of the glutamate receptor [NMDA], precursor of the zeta I subunit of the glutamate receptor [NMDA], precursor of the alpha-1 chain of the glycine receptor, precursor of the glycine receptor alpha-2, glycine receptor alpha-3 chain precursor, glycine receptor beta chain precursor, ribonucleoprotein H/ACA complex subunit 1, high affinity immunoglobulin epsilon receptor beta subunit , hypothetical protein DKFZp313I0334, hypothetical protein DKFZp761M1724, hypothetical protein FLJ12242, hypothetical protein FLJ14389, hypothetical protein FLJ14798, hypothetical protein FLJ14995, hypothetical protein FLJ16180, hypothetical protein FLJ16802, hypothetical protein FLJ32069, hypothetical FLJ37401, hypothetical protein FLJ38750, hypothetical protein FLJ40162, protein hypothetical FLJ41415, hypothetical protein FLJ90576, hypothetical protein FLJ90590, hypothetical protein FLJ90622, hypothetical protein KCTD15, hypothetical protein MGC15619, inositol 1,4,5-trisphosphate receptor type 1, inositol 1,4,5-trisphosphate receptor type 2 , inositol 1,4,5-trisphosphate receptor type 3, intermediate-conductance calcium-activated potassium channel protein 4, inwardly rectifying potassium channel 13, inwardly rectifying potassium channel 16, inwardly rectifying potassium channel 4, Kir2.2v negative regulator of inwardly rectifying K(+) channels, KA2a subunit of the kainate receptor, KCNH5 protein, KCTD protein 17, KCTD2 protein, keratinocyte-associated transmembrane protein 1, Kv channel-interacting protein 4, melastatin 1, membrane protein MLC1, MGC15619 protein, mucolipin-1, mucolipin-2, mucolipin-3, multidrug resistance-associated protein 4, N-methyl-D-aspartate receptor subunit 2C precursor, homolog of NADPH oxidase 1, neuronal acetylcholine receptor protein Nav1.5, alpha-10 subunit precursor, neuronal acetylcholine receptor protein, alpha-2 subunit precursor, neuronal acetylcholine receptor protein, alpha-subunit precursor 3, neuronal acetylcholine receptor protein, alpha-4 subunit precursor, neuronal acetylcholine receptor protein, alpha-5 subunit precursor, neuronal acetylcholine receptor protein, alpha-6 subunit precursor, receptor protein neuronal acetylcholine, alpha-7 subunit precursor, neuronal acetylcholine receptor protein, alpha-9 subunit precursor, neuronal acetylcholine receptor protein, beta-2 subunit precursor, neuronal acetylcholine receptor protein, precursor beta-3 subunit, neuronal acetylcholine receptor protein, beta-4 subunit precursor, 2D alpha subunit of neuronal voltage-gated calcium channels, purinergic P2X receptor 1, purinergic P2X receptor 2, purinergic P2X receptor 3, receptor purinergic P2X 4, purinergic P2X receptor 5, purinergic P2X receptor 6, purinergic P2X receptor 7, pancreatic potassium channel TALK-1b, pancreatic potassium channel TALK-1c, pancreatic potassium channel TALK-1d, Phospholemman precursor, plasmolipin, protein polycystic kidney disease-related 2, polycystic kidney disease-like protein 2, polycystic kidney disease-like protein 2, polycystic kidney disease and receptor for egg gelatin-related protein precursor, polycystin -2, potassium channel regulator, potassium channel K subfamily member 1, potassium channel K subfamily member 10, potassium channel K subfamily member 12, potassium channel K subfamily member 13 potassium channels, potassium channel K subfamily member 15, potassium channel K subfamily member 16, potassium channel K subfamily member 17, potassium channel K subfamily member 2, member 3 potassium channel subfamily K, potassium channel K subfamily member 4, potassium channel K subfamily member 5, potassium channel K subfamily member 6, channel K subfamily member 7 potassium, potassium channel subfamily K member 9, potassium channel tetramerization domain-containing protein 12, potassium channel tetramerization domain-containing protein 14, protein 2 containing potassium channel tetramerization domains, protein 4 containing potassium channel tetramerization domains, protein 5 containing potassium channel tetramerization domains, 10 containing potassium channel tetramerization domains, protein 13 containing potassium channel tetramerization domains, 1 containing potassium channel tetramerization domains, member 1 of potassium voltage-gated channel subfamily A, member 2 of potassium voltage-gated channel subfamily A, member 4 of potassium voltage-gated channel subfamily A, potassium voltage-gated channel subfamily A member 5, potassium voltage-gated channel subfamily A member 6, voltage-gated channel subfamily B member 1 potassium, potassium voltage-gated channel subfamily B member 2, potassium voltage-gated channel subfamily C member 1, potassium voltage-gated channel subfamily C member 3, potassium voltage-gated channel subfamily C member 4 Potassium voltage-gated channel C, Potassium voltage-gated channel subfamily D member 1, Potassium voltage-gated channel subfamily D member 2, Potassium voltage-gated channel subfamily D member 3 , member 1 of the E subfamily of potassium voltage-gated channels, member 2 of the E subfamily of potassium voltage-gated channels, member 3 of the E subfamily of potassium voltage-gated channels, member 4 of the E subfamily of potassium voltage-gated channels, member 1 of the F subfamily of potassium voltage-gated channels, member 1 of the G subfamily of potassium voltage-gated channels, member 2 of the G subfamily of potassium voltage-gated channels, member Potassium voltage-gated channel subfamily G 3, potassium voltage-gated channel subfamily G member 4, potassium voltage-gated channel subfamily H member 1, channel subfamily H member 2 potassium voltage-gated channel subfamily H member 3, potassium voltage-gated channel subfamily H member 4, potassium voltage-gated channel subfamily H member 5, member 6 of the H subfamily of potassium voltage-gated channels, member 7 of the H subfamily of potassium voltage-gated channels, member 8 of the H subfamily of potassium voltage-gated channels, member 1 of the KQT subfamily of potassium voltage-gated channels potassium voltage, member 2 of the KQT subfamily of potassium voltage-gated channels, member 3 of the KQT subfamily of potassium voltage-gated channels, member 4 of the KQT subfamily of potassium voltage-gated channels, member 5 of the KQT subfamily of potassium voltage-gated channels, member 1 of the S subfamily of potassium voltage-gated channels, member 2 of the S subfamily of potassium voltage-gated channels, member 3 of the S subfamily of potassium voltage-gated channels potassium, potassium voltage-gated channel subfamily V member 2, potassium voltage-gated channels, subfamily H, member 7, isoform 2, potassium/sodium hyperpolarization-activated cyclic nucleotide-gated channel 1, nucleotide-gated channel potassium/sodium hyperpolarization-activated cyclic 2, potassium/sodium hyperpolarization-activated cyclic nucleotide-gated channel 3, potassium/sodium hyperpolarization-activated cyclic nucleotide-gated channel 4, probable mitochondrial import receptor subunit TOM40 homologs , purinergic receptor, isoform A, putative 4-repeat voltage-selected ion channel, putative chloride channel protein 7, putative GluR6 kainate receptor, putative ion channel protein CATSPER2 variant 1, CATSPER2 protein variant 2 putative ion channel protein, putative ion channel protein CATSPER2 variant 3, putative regulator of potassium channel protein variant 1, putative TPTE protein tyrosine phosphatase, ryanodine receptor 1, ryanodine receptor 2, ryanodine receptor 3 ryanodine, SH3KBP1 binding protein 1, short transient receptor potential channel 1, short transient receptor potential channel 4, short transient receptor potential channel 5, short transient receptor potential channel 6, short transient receptor potential channel 7, protein small-conductance calcium-activated potassium channel 1, small-conductance calcium-activated potassium channel protein 2, isoform b, small-conductance calcium-activated potassium channel protein 3, isoform b, calcium-activated potassium channel SK2 small-conductance calcium-activated potassium channel, SK3 small-conductance calcium-activated potassium channel, sodium channel, sodium channel beta-1 subunit precursor, sodium channel protein type II alpha subunit, sodium channel protein type III alpha subunit sodium channels, sodium channel protein type IV alpha subunit, sodium channel protein type IX alpha subunit, sodium channel protein type V alpha subunit, sodium channel protein type VI alpha subunit, alpha subunit sodium channel protein type VIII, sodium channel protein type /potassium, sperm-associated cation channel 1, sperm-associated cation channel 2, isoform 4, syntaxin-1 B1, transient receptor potential cation channel subfamily A member 1, channel subfamily M member 2 receptor potential cation channels, member 3 of the M subfamily of transient receptor potential cation channels, member 6 of the M subfamily of transient receptor potential cation channels, member 7 of the M subfamily of receptor potential cation channels transient, member 1 of the subfamily V of transient receptor potential cation channels, member 2 of the subfamily V of transient receptor potential cation channels, member 3 of the subfamily V of transient receptor potential cation channels, member 4 of transient receptor potential cation channel subfamily V, member 5 of transient receptor potential cation channel subfamily V, member 6 of transient receptor potential cation channel subfamily A, epsilon splice variant of channels transient receptor potential 4, zeta splice variant of transient receptor potential 4 channels, gamma splice variant of transient receptor potential 7 channels, tumor necrosis factor, protein 1 alpha-induced, endothelial, protein 2 two-pore calcium channels, VDAC4 protein, voltage-gated potassium channel Kv3.2b, voltage-gated sodium channel beta1B subunit, voltage-gated anion channels, voltage-gated anion channels 2, selective channel protein 1 voltage-gated anion channel protein 2, voltage-gated anion-selective channel protein 3, voltage-gated calcium channel gamma-1 subunit, voltage-gated calcium channel gamma-2 subunit voltage, gamma-3 subunit of voltage-gated calcium channels, gamma-4 subunit of voltage-gated calcium channels, gamma-5 subunit of voltage-gated calcium channels, gamma-6 subunit of voltage-gated calcium channels, gamma-7 subunit of voltage-gated calcium channels, gamma-8 subunit of voltage-gated calcium channels, alpha-1 C subunit of voltage-gated L-type calcium channels, alpha-1 D subunit of L-type calcium channels voltage-gated, alpha-1S subunit of voltage-gated L-type calcium channels, beta-1 subunit of voltage-gated L-type calcium channels, beta-2 subunit of voltage-gated L-type calcium channels, beta-3 subunit of voltage-gated L-type calcium channels, beta-4 subunit of voltage-gated L-type calcium channels, alpha-1B subunit of voltage-gated N-type calcium channels, alpha-1A subunit of P/G-type calcium channels voltage-gated, alpha-1E subunit of voltage-gated R-type calcium channels, alpha-1G subunit of voltage-gated T-type calcium channels, alpha-1H subunit of voltage-gated L-type calcium channels, alpha-11 subunit of voltage-gated T-type calcium channels, alpha-1 subunit of voltage-gated L-type calcium channels, beta-1 subunit of voltage-gated potassium channels, beta-2 subunit of voltage-gated potassium channels, beta subunit -3 voltage-gated potassium channels, KCNA7 voltage-gated potassium channel.

[0213]Exemplary GPCRs include, but are not limited to, class A rhodopsin like receptors such as vertebrate muscle acetylcholine type 1, vertebrate muscle acetylcholine type 2, vertebrate muscle acetylcholine type 3, vertebrate muscle acetylcholine type 4; adrenergic receptors (alpha type 1 adrenergic receptors, alpha type 2 adrenergic receptors, beta type 1 adrenergic receptors, beta type 2 adrenergic receptors, beta type 3 adrenergic receptors, vertebrate dopamine type 1, vertebrate dopamine type 2, vertebrate dopamine type 3 , vertebrate dopamine type 4, histamine type 1, histamine type 2, histamine type 3, histamine type 4, serotonin type 1, serotonin type 2, serotonin type 3, serotonin type 4, serotonin type 5, serotonin type 6, serotonin type 7 , serotonin type 8, other types of serotonin, Trace amine, angiotensin type 1, angiotensin type 2, bombesin, bradykinin, anaphylatoxin C5a, Fmet-leu-phe, APJ-like, interleukin-8 type A, interleukin-8 type B, other types of interleukin-8, chemokine C-C type 1 to type 11 and other types, chemokine C-X-C (types 2 to 6 and others), chemokine C-X3-C, cholecystokinin CCK, CCK type A, CCK type B, other CCK, endothelin, melanocortin (melanocyte-stimulating hormone, adrenocorticotropic hormone, melanocortin hormone), Duffy antigen, prolactin-releasing peptide (GPR10), neuropeptide Y (type 1 to 7), neuropeptide Y, other neuropeptides Y, neurotensin, opioid (type D, K, M, 1 to 2), vasotocin, oxytocin/mesotocin, conopressin, galanin-like, activated proteinase-like, orexin and FF neuropeptides, QRFP, chemokine receptor-like, neuromedin U-like (neuromedin U, PRXamide), hormone protein ( follicle-stimulating hormone, lutropin-choriogonadotropic hormone, thyrotropin, gonadotropin type I, gonadotropin type II), (rod)opsin, vertebrate rhodopsin (types 1-5), vertebrate rhodopsin type 5, arthropod rhodopsin, arthropod rhodopsin type 1, arthropod rhodopsin type 2, arthropod rhodopsin type 3, mollusk rhodopsin, rhodopsin, olfactory (olfactory II fam 1 to 13), prostaglandin (prostaglandin E2 subtype EP1, prostaglandin E2/D2 subtype EP2, prostaglandin E2 subtype EP3, prostaglandin E2 subtype EP4, prostaglandin F2-alpha, prostacyclin, thromboxane, adenosine type 1 to 3, purinergic receptors, purinergic receptor P2RY1-4, 6, 11 GPR91, purinergic receptor P2RY5, 8, 9, 10 GPR35, 92, 174, purinergic receptor P2RY12-14 GPR87 (UDP-Glucose), cannabinoid, platelet activating factor, gonadotropin-releasing hormone, gonadotropin-releasing hormone type I, gonadotropin-releasing hormone type II, adipokinetic hormone-like, corazonin, thyrotropin-releasing hormone and secretagogue , thyrotropin-releasing hormone, growth hormone secretagogue, growth hormone secretagogue-like, ecdysis-triggering hormone (ETHR), melatonin, lysosphingolipid and LPA (EDG), sphingosine 1-phosphate Edg-1, lysophosphatidic acid Edg-2 , sphingosine 1-phosphate Edg-3, lysophosphatidic acid Edg-4, sphingosine 1-phosphate Edg-5, sphingosine 1-phosphate Edg-6, lysophosphatidic acid Edg-7, sphingosine 1-phosphate Edg-8, Edg. Leukotriene B4 receptor G protein-coupled bile acid, free fatty acid receptor (GP40, GP41, GP43), class B secretin-like, calcitonin, corticotropin-releasing factor, gastric inhibitory peptide, glucagon, growth hormone-releasing hormone, parathyroid hormone, PACAP, secretin, vasoactive intestinal polypeptide, latrophilin, latrophilin type 1, latrophilin type 2, latrophilin type 3, ETL receptors, brain-specific angiogenesis inhibitor (BAI), Methuselah-like proteins (MTH), EGF LAG cadherin (CELSR) , very large G protein-coupled receptor, class C metabotropic glutamate/pheromone, group I to III metabotropic glutamate, calcium sensitization-like, extracellular calcium sensitization, pheromone, other calcium sensitization-like, putative pheromone receptors, GABA -B, GABA-B subtype 1, GABA-B subtype 2, GABA-B-like, Orphan GPRC5, Orphan GPCR6, Bride of sevenless proteins (BOSS), Taste receptors (T1R), Class D fungal pheromones, similar to fungal pheromone factor A (STE2, STE3), fungal pheromone B-like (BAR, BBR, RCB, PRA), cAMP receptors class E, ocular albinism proteins, Frizzled/Smoothened family, frizzled group A (F z 1 and 2 and 4 and 5 and 7-9), frizzled group B (Fz 3 and 6), frizzled group C (others), vomeronasal receptors, nematode chemoreceptors, insect odorant receptors and archaeal/bacterial/fungal class opsins Z.

[0214]The target MURPs can be designed to target any cellular protein including, but not limited to, cell surface protein, secretase protein, cytosolic protein, and nuclear protein. One target of particular interest is an ion channel.

[0215]Ion channels constitute a superfamily of proteins, including the potassium channel family (K channels), the sodium channel family (Na channels), the calcium channel family (Ca channels) , the chloride channel family (Cl channels) and the acetylcholine channel family. Each of these families contains subfamilies and each subfamily typically contains specific channels derived from individual genes. For example, the K channel family contains subfamilies of voltage-gated potassium channels called Kv1.x and Kv3.x. The Kv1.x subfamily contains the channels Kv1.1, Kv1.2 and Kv1.3, which correspond to the products of individual genes and are thus called 'species'. The classification applies to the Na, Ca, Cl channel families and other families as well.

[0216]Ion channels can also be classified based on the mechanisms by which the channels operate. Specifically, the major types of ion channel proteins are characterized by the procedure used to open or close the channel protein to both allow and prevent specific ions from entering the channel protein and passing through a lipid bilayer cell membrane. An important type of channel protein is the voltage-gated channel protein that opens or closes (regulates) in response to changes in the electrical potential across the cell membrane. Voltage-gated sodium channel 1.6 (Nav1.6) is of particular interest as a therapeutic target. Another type of ion channel protein is the mechanically gated channel, for which a mechanical stress on the protein opens or closes the channel. Still another type is called a ligand-gated channel, which opens or closes depending on whether or not a particular ligand binds to the protein. The ligand may be either an extracellular moiety, such as a neurotransmitter, or an intracellular moiety, such as an ion or nucleotide.

[0217]Ion channels generally allow passive flow of ions toward an electrochemical gradient, while ion pumps use ATP to transport against a gradient. Coupled transporters, both antiporters and symporters, allow the movement of one ion species against its gradient, driven by the downward movement of another ion species.

[0218]One of the most common types of channel proteins, found in the membrane of almost all animal cells, allows the specific permeation of potassium ions across a cell membrane. In particular, potassium ions enter rapidly through cell membranes to K+ channel proteins (up to 10-8 ions per second). Furthermore, potassium channel proteins have the ability to distinguish between potassium ions and other small alkali metal ions such as Li+ or Na+ with high fidelity. In particular, potassium ions are at least ten thousand times more permanent than sodium ions. Potassium channel proteins typically comprise four subunits (usually identical), thus their cell surface targets are present as tetramers, allowing tetravalent binding of MURP. One type of subunit contains six hydrophobic segments in length (which can span the membrane), while the other types contain two hydrophobic segments.

[0219]Another significant family of channels is the calcium channel. Calcium channels are generally classified according to their electrophysiological properties as low-voltage-activated (LVA) or high-voltage-activated (HVA) channels. HVA channels comprise at least three groups of channels, known as type L, N and P/Q channels. These channels have been differentiated from each other electrophysiologically, as well as biochemically, based on their pharmacology and ligand binding properties. For example, dihydropyridines, diphenylalkylamines and piperidines bind to the AI subunit of the L-type calcium channel and block a proportion of HVA calcium currents in neuronal tissue, which are called L-type calcium currents. N-type calcium channels are sensitive to omega conopeptides, but are relatively insensitive to dihydropyridine compounds such as nimodipine and nifedipine. P/Q-type channels, on the other hand, are insensitive to dihydropyridines, but are sensitive to the funnel-web spider toxin Aga III<a>. R-type calcium channels, like L-, N-, P-, and Q-type channels, are activated by large membrane depolarizations, and are thus classified as high-voltage-activated (HVA) channels. R-type channels are generally insensitive to dihydropyridines and omega conopeptides, but, like P/Q, L, and N channels, they are sensitive to the funnel-web spider toxin AgaIVA. Immunocytochemical staining studies indicate that these channels are located throughout the brain, particularly in the deep midline structures (caudate-putamen, thalamus, hypothalamus, amygdala, cerebellum) and in the nuclei of the ventral midbrain and brainstem. Neuron voltage-sensitive calcium channels! They typically consist of a core ai subunit, an a<2>/S subunit, a p subunit, and a 95 kD subunit.

[0220]Additional non-limiting examples include Kir (an internally rectifying potassium channel), Kv (a voltage-gated potassium channel), Nav (a voltage-gated sodium channel), Cav (a voltage-gated calcium channel ), CNG (cyclic nucleotide-gated channel), HCN (hyperpolarization-activated channel), TRP (a transient receptor potential channel), CIC (a chloride channel), CFTR (a cystic fibrosis transmembrane conductance regulator, a chloride channel), IP3R (an inositol trisphosphate receptor), RYR (a ryanodine receptor). Other types of channels are 2-pore channels, glutamate receptors (AMPA, NMDA, KA), M2, connexins, and Cys-loop receptors.

[0221]A common design for ion channel proteins such as Kv1.2, Kv3.1, Shaker, TRPC1 and TRPC5 is to have six membrane-spanning segments arranged as follows:

N end-S1-E1-S2-X1-S3-E2-S4-X2-S5-E3-S6-C end

[0222]Wherein S1-6 are membrane-spanning sequences, E1-3 are extracellular surface loops, and X1-2 are intracellular surface loops. The E3 loop is generally the longest of the three extracellular loops and is hydrophilic so it is a good target for drug and MURP binding. The pore-forming part of most channels is a multimeric (e.g., tetrameric or rarely pentameric) complex of alpha helices that spans the membrane. There is usually a pore loop, which is a region of the protein that returns to the membrane to form the selectivity filter that determines which ion species can enter. Such channels are called 'pore-loop' channels.

[0223]Ion channels are valuable targets for drug design because they participate in a wide range of physiological processes. In humans there are approximately more than three hundred ion channel proteins, many of which are involved in genetic diseases. For example, abnormal expression or function of ion channels has been shown to result in a wide range of diseases including cardiac, neuronal, muscle, respiratory, and metabolic diseases. This section is based on ion channels, but the same concepts and approaches are equally applicable to all membrane proteins, including 7TM, 1TM, G proteins and G protein-coupled receptors (GPCRs), etc. Some of the ion channels are GPGR.

[0224]Ion channels typically form large macromolecular complexes that include closely linked accessory protein subunits, and the combinatorial use of such subunits contributes to the diversity of ion channels. These accessory proteins may also be the binding targets of the target MURPs, microproteins and toxins.

[0225]The subject MURPs may be designed to bind to any of the channels known in the art and those specifically exemplified herein. MURPs exhibiting a desired ion channel binding capacity (encompassing specificity and avidity) can be selected by any recombinant and biochemical techniques (e.g., expression and display) known in the art. For example, MURPs can be displayed by a genetic package including, but not limited to, phages and spores, and subjected to immunopurification against intact cell membranes, or preferably intact cells such as whole mammalian cells. To eliminate phage binding to other non-target cell surface molecules, the conventional approach was to perform subtraction immunopurification against similar cell lines that had a low or no detectable level of the target receptor. However, Popkov et al. (J. Immunol. Methods 291:137-151 (2004)) showed that related cell types are not ideal for subtraction because they generally have a reduced, but still significant, level of the target on their surface, which reduces the number of desired phage clones. This problem occurs even when immunopurifying on cells that have been transfected with the gene encoding the target, followed by negative selection/subtraction on the same cell line that was not transfected, especially if the native target gene was not inactivated. Instead, Popkov et al. showed that negative subtraction selection or immunopurification works much better if performed with an excess of the same cells used for normal immunopurification (positive selection), except that the target has now been blocked with a high-affinity target-specific inhibitor. such as a small molecule, peptide or an antibody to the target, which makes the active site unavailable. This procedure is called “negative selection with epitope-masked cells,” which is particularly useful in selecting MURP targets with a desired ion channel binding capacity.

[0226] Microproteins are also disclosed herein, and particularly microproteins that have binding capacity towards at least one family of ion channels. A genetic package that visualizes such microproteins is also disclosed in the present document. Non-limiting examples of ion channels to which the target microproteins bind are sodium, potassium, calcium, acetylcholine and chloride channels. Of particular interest are those microproteins and the genetic packages that display such microproteins, which have binding capacity towards native targets. Native targets are generally natural molecules or fragments derived therefrom that the microprotein is known to bind, typically including those known binding targets that have been reported in the literature.

[0227] Also disclosed herein is a genetic package that displays an ion channel binding microprotein that has been modified. The modified microprotein may (a) bind to a different family of channel relative to the corresponding unmodified microprotein; (b) binding to a different subfamily of the same channel family relative to the corresponding unmodified microprotein; (c) binding to a different species of the same channel subfamily with respect to the corresponding unmodified microprotein; (d) the microprotein binds to a different site on the same channel relative to the corresponding unmodified microprotein; and/or (e) bind to the same site of the same channel, but give a different biological effect with respect to the corresponding unmodified microprotein.

[0228]Figure 22 and 46 show how microprotein or toxin domains that each bind to different sites of the same ion channel can be combined into a single protein. The two binding sites where these two microproteins bind can be on two channels from different families, two channels from the same family, but a different subfamily, two channels from the same subfamily, but a different species (gene product), o two different binding sites on the same channel (species) or they can bind (simultaneously or not) to the same binding site on the same channel (species), since the channels are multimeric. Binding modules and domains that bind to sites on the channels can be microprotein domains (natural or non-natural, containing 2 to 8 disulfides), one-disulfide peptides, or linear peptides. These modules can be selected independently and combined, or they can be selected from a library for joining in the presence of a fixed active joining module. In the latter case, the display library would display multiple modules of which one would contain a variant library. A typical goal is to select a dimer from this library that has a higher affinity than the active monomer that was the starting point.

[0229] Also disclosed herein is a protein comprising a plurality of ion channel binding domains, wherein the individual domains are microprotein domains that have been modified such that.

(a) the microprotein domains bind a different channel family relative to the corresponding unmodified microprotein domains; (b) the microprotein domains bind to a different subfamily of the same channel family relative to the corresponding unmodified microprotein domains; (c) the microprotein domains bind to a different species of the same subfamily relative to the corresponding unmodified microprotein domains; (d) the microprotein domains bind to a different site on the same channel relative to the corresponding unmodified microprotein domains; (e) the microprotein domains bind to the same site of the same channel, but give a different biological effect relative to the corresponding unmodified microprotein domains; and/or (f) the microprotein domains bind to the same site of the same channel and give the same biological effect with respect to the corresponding unmodified microprotein domains. If desired, the microprotein domains may comprise natural or non-natural sequences. The individual domains can be linked together by a heterologous linker. Individual microprotein domains may bind to the same or different channel family, the same or different channel subfamily, the same or different species of the same subfamily, the same or different site on the same channel.

[0230]The target microproteins may be a toxin. Preferably, the toxin retains part or all of its toxicity spectrum. In particular, venomous animals, such as snakes, have a range of prey and intruder species and venom toxins differ in activity for different receptors of different species. Venom consists of a large number of related and unrelated toxins, each toxin having a “spectrum of activity,” which can be defined as all the receptors of all the species on which that toxin has measurable activity. All targets in the 'activity spectrum' are considered “native targets” and this includes any human target against which the toxin is active. The native target(s) of a microprotein or toxin include all targets reported to be inhibited by the toxin in the literature. The higher the affinity or activity on a target, the more likely that target is the natural native target, but it is common for toxins that act on multiple targets within the same species. The native target(s) may be human or non-human receptor(s) against which the toxin is active.

[0231]For the toxin to retain the ability to bind to cells after fusion with the display vector, it may be desired to test both the N-terminus and the C-terminus for fusion and to test a variety of fusion sites (i.e. 0, 1, 2, 3, 4, 5, 6 amino acids before the first cysteine or after the last cysteine of the toxin domain, if the toxin domain is a cysteine-containing domain) using a synthetic DNA library approach , which preferentially encodes a library of glycine-rich linkers, which form the smallest amino acid chain, are uncharged, and are most likely compatible with the binding of the toxin to the target. Since the N-terminal amino group and C-terminal carboxyl groups can participate in target binding, the library should contain a lysine or an arginine to mimic the positively charged amino group (or fusions to the N-terminus of the toxin) and a glutamate or an aspartate to mimic the negatively charged carboxyl group (for fusions to the C terminus of the toxin).

[0232]The inhibitor(s) used to block the target during negative selection can be small molecules, peptides or proteins, and natural or unnatural. In addition to simple subtraction, the choice of inhibitor mixture is a valuable tool to control the specificity of the ion channel inhibitors being designed. Because there are more than three hundred ion channels in total, with partially overlapping specificities and sequence similarities, and multiple modulatory sites per channel, each having a different effect, the specificity requirement can be complex.

[0233]If the activity of a toxin is modified, or if two different toxins are combined into a single protein, the two toxins can bind to the same channel at the same site and have the same physiological effect, or the two toxins can bind to the same channel at the same site and have a different physiological effect, or the two toxins may bind to the same channel at a different site, or the two toxins may bind to different channels that belong to the same subfamily (i.e., Kv1.3 and Kv1.2; meaning product of a different gene or 'species'), or the two toxins can bind to different channels that belong to the same family (i.e., they are both K channels), or the two toxins can bind to channels that belong to different families (i.e., K channels vs. Na channels).

[0234]Ion channels typically have many transmembrane segments (24 for sodium channels) and thus offer several different, non-competitive, non-overlapping binding sites for modulators to alter channel activity in different ways. One approach is to create binders for one site on the same ion channel as existing binders for a different site, even if these sites are unrelated. To achieve this, the existing toxin can be used as a targeting agent for a library of 1, 2, 3, or 4 disulfide proteins that are separated from the targeting toxin by a flexible 5, 6, 7, 8, 9, 10, 12, 14, 16, 18, 20, 25, 30, 35, 40 or 50 amino acids. It is useful if the affinity of the targeting agent is not too high, so that the affinity of the new library can have a significant contribution to the overall affinity. Another approach is to create new modulators for channels from existing modulators for other channels that are related in sequence or structure. The conotoxin family, for example, contains sequence-related and structure-related modulators for Ca, K, Na channels, and nicotinic acetylcholine receptors. It seems feasible to convert a K channel modulator to a Na channel modulator using a library of conotoxin derivatives, or vice versa. For example, kappa conotoxins inhibit K channels, mu conotoxins and delta conotoxins inhibit Na channels, omega conotoxins inhibit Ca channels, and alpha conotoxins inhibit acetylcholine receptors.

[0235]The proximity of different binding sites, each with a different effect on channel activity, of the same ion channel makes it attractive to ligate inhibitors using flexible linkers, which create a single inhibitor with two domains, each binding in a different site, or a single protein with two domains that bind on different copies of the same site, giving a high-affinity bivalent interaction (avidity). This approach has not been taken for natural toxins, supposedly because they must act quickly and thus remain small in order to have maximum tissue penetration, but for pharmaceuticals the speed of action is less important, making this an attractive approach.

[0236]Combinatorial libraries of dimeric, trimeric, tetrameric or multimeric toxins/modulators, each native or modified, can thus be created and directly screen these libraries at the protein level or immunopurify these libraries using gene packages for improved affinity (avidity, if produces binding simultaneously at multiple sites) and then characterize the specificity and activity of such multimeric clones by protein expression and purification, followed by cell-based activity assays, including patch binding assays. Individual modules can be immunopurified and selected separately, in isolation from each other, or they can be engineered in the presence of each other, so that the new domain is added to a display system as a library that also contains a fixed active copy that serves as targeting element for the library and only clones that are significantly better than the fixed active monomer are selected and characterized.

[0237]Figs. 46 and 47 show some of the monomeric derivatives that can be prepared from native (natural) proteins, and some of the multimers that can be prepared to bind to multiple different binding sites on the target. The linkers are shown as glycine-rich PEGr, but the linkers could be any sequence and could also be optimized using molecular libraries, followed by immunopurification. Libraries can be created within the active native toxin itself using a variety of mutagenesis strategies as described above, or they can extend the existing area of contact with the target by creating libraries on the N-terminal or C-terminal side of the active toxins, hoping to create additional contacts with the target. Such libraries may be based on existing toxins with known activity for that site, or may be naive 1, 2, 3, 4 disulfide libraries based on unrelated microprotein scaffolds. These additional contact elements can be added to one or both sides of the active domains, and can be directly adjacent to the existing modulatory domain or can be separated from it by flexible linkers. The initial multimer or the final improved multimer may be a homomultimer or a heteromultimer, based on sequence similarity of the domains or based on the target specificity of the domains of the multimer. Thus, the monomers comprising the multimer may bind to the same target sites, but have the same or different sequences. With 10-100 different native toxins that are known to bind to each channel family, and with 2, 3, 4, 5 or 6 domains per clone, display libraries can be created with enormous combinatorial diversity, although only one uses sequences of native toxin. Low-level synthetic mutagenesis based on amino acid similarity or phylogenetic substitution rates within the family can be used to create high-quality libraries of mutants, of which a very high fraction is expected to retain function, with a high probability enhanced function in some of the properties of interest.

[0238]The binding capacity of target MURPs, microproteins or toxins to a given ion channel can be measured in terms of the Hill coefficient. The Hill coefficient indicates the stoichiometry of the binding interaction. A Hill coefficient of 2 indicates that 2 inhibitors bind to each channel. Allosteric modulation, which is the modulation of activity at one site produced by binding at a remote site, can also be assessed.

[0239]The biological activity or effect of an ion channel and the ability of the target MURP, microproteins or toxins to regulate an ion channel activity can be evaluated using a variety of in vitro and in vivo assays. For example, available in the art procedures for measuring voltage, measuring current, measuring membrane potential, measuring ion flux, for example, potassium or rubidium, measuring ion concentration, measuring regulation, measuring second messengers and transcription levels, and using, for example, color-sensitive dyes voltage, radioactive tracers and patch fixation electrophysiology. In particular, such assays can be used to test microproteins and toxins that can inhibit or activate an ion channel of interest.

[0240]Specifically, potential channel inhibitors or activators can be tested in comparison to an appropriate control to examine the degree of modulation. Control samples may also be samples untreated with the candidate activators or inhibitors. Inhibition is present when a given ion channel activity value relative to the control is approximately 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, or even less. The IC<50> is a commonly used unit (the concentration of inhibitor that reduces ion channel activity by 50%) to determine the inhibitory effect. Similar to IQ<90>. Channel activation is achieved when a given ion channel activity value relative to control increases by 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100 %, 200%, 500%, or more.

[0241]Changes in ion flux can be assessed by determining changes in the polarization (i.e., electrical potential) of the cell or membrane expressing the channel of interest. For example, one procedure is to determine changes in cell polarization by measuring changes in current (thus measuring changes in polarization) with voltage clamping and patch clamping techniques, for example, the “cell-attached” mode, the “ the other way around” and the “whole cell” mode (see, for example, Ackerman et al., New Engl. J. Med. 336:1575-1595 (1997)). Whole cell currents are conveniently determined using conventional methodology (see, for example, Hamil et al., Pflugers. Archiv. 391:85 (1981). Other known assays include: radiolabeled rubidium flux assays and fluorescence assays using voltage-sensitive dyes (see, for example, Vestergarrd-Bogind et al., J. Membrane Biol. 88:67-75 (1988); Daniel et al., J. Pharmacol. Meth. 25:185-193 (1991) ; Holevinsky et al., J. Membrane Biology 137:59-70 (1994)).

[0242]The effects of candidate MURPs, microproteins or toxins on the function of a channel of interest can be measured by changes in electrical currents or ion flow or by the consequences of changes in currents and flow. The downstream effect of candidate proteins on ion flux can be varied. Therefore, any suitable physiological change can be used to evaluate the influence of a candidate protein on the test channels. The effects of the candidate protein can be measured by a toxin binding assay. If functional consequences are determined using intact cells or animals, a variety of effects can also be measured such as transmitter release (e.g. dopamine), hormone release (e.g. insulin), transcriptional changes to both known and unknown genetic markers. characterized (e.g., Northern blots), cell volume changes (e.g., in red blood cells), immune responses (e.g., T cell activation), changes in cell metabolism such as cell growth or pH changes, and changes in intracellular second messengers such as Ca2+.

[0243]Other key biological activities of ion channels are ion selectivity and regulation. Selectivity is the ability of some channels to discriminate between ion species, allowing some to pass into the pore while excluding others. Regulation is the transition between open and closed states. It can be evaluated by any of the procedures known in the art or disclosed herein[0244]Still another biological property by which the target MURP, microprotein or toxin can be selected is the frequency of opening and closing of the target channels, called frequency of regulation. The regulation frequency is influenced by voltage (in voltage-gated channels, which open or close by changes in membrane voltage) and ligand binding. The transition rate between open and closed states is typically <10 microseconds, but can be increased or decreased by other molecules. The flow rate (current) in the pore when it opens is on the order of 10e7 ions per second for ion channels and much less for coupled exchangers. Upon opening, some voltage-gated channels enter an inactivated nonconducting state in which they are resistant to depolarization.

EXAMPLES

Example: Design of an allicin-serine oliaomer based on human sequences

[0245]The human genome database was searched for sequences that were rich in glycine. Three sequences were identified as suitable donor sequences as shown in Table X.

Table X: Donor sequences for GRS design A.

Access Protein Amino Acid Sequences

NP_009060 GGGGSGGGSGSGGGG 486-499 zinc finger protein

Q9Y2X9 GSGSGGGGSGG 19-31 zinc finger protein CAG38801 SGGGGSGGGSGSG 7-19 MAP2K4

[0246]Based on the sequences in Table Peptide A can oligomerize to form structures with the formula (GGGSGSGGGGS)n in which n is between 2 and 40. Figure 5 shows that all possible 9-mer subsequences in oligomers of peptide A are contained in at least one of the proteins listed in Table 3. Thus, peptide A oligomers do not contain human T cell epitopes. Inspection of Figure 5 reveals that GRS based on peptide A oligomers can start and end at any of the positions of peptide A.

Example: Design of a glycine-proline oligomer based on human sequences

[0247]Glycine-rich sequences were designed based on the sequence GPGGGGGPGGGGGPGGGGPGGGGGGGPGGGGGGPGGG representing amino acids 146-182 of the human class 4 POU domain with accession number NP_006228. Figure 6 illustrates that peptide B oligomers with the sequence GGGGGPGGGGP can be used as GRS. All 9-mer subsequences that are contained in peptides with the sequence (GGGGPGGGGP)n are also contained in the POU domain sequence. Thus, such oligomeric sequences do not contain T cell epitopes.

[0248]

Example: Design of alicine-glutamic acid oligomer

[0249]Glycine-rich sequences can be designed based on the GAGGEGGGGEGGGPGG subsequence which is part of the ribosomal protein S6 kinase (accession number BAD92170). For example, C-peptide oligomers with the sequence GGGGE will form sequences in which the majority of the 9-mer subsequences will be contained in the S6 ribosomal protein kinase sequence. Thus, oligomeric GRSs of the general structure (GGGGE)n have a very low risk of containing T cell epitopes.

Example: Identification of human hydrophilic glycine-rich sequences

[0250]A human protein database was searched for subsequences that were rich in glycine residues. These subsequences contained at least 50% glycine. Only the following non-glycine residues were allowed to occur in the GRS: ADEHKPRST. 70 subsequences were identified that had a minimum length of 20 amino acids. These subsequences are listed in Appendix A. They can be used to construct GRSs with low immunogenic potential in humans.

Example: Construction of rPEG J288

[0251]The following example describes the construction of a codon-optimized gene that encodes a URP sequence with 288 amino acids and the sequence (GSGGEG)48. First, the present inventors constructed a filler vector pCW0051 as illustrated in Fig. 40. The sequence of the expression cassette in pCW0051 is shown in Fig. 42. The filler vector was based on a pET vector and included a promoter T7. The vector encodes a Flag sequence, followed by a filler sequence that is flanked by BsaI, BbsI, and KpnI sites. The BsaI and BbsI sites were inserted in a way that generated compatible overhang residues after digestion as illustrated in Fig. 42. The filler sequence was followed by a His6 tag and the green fluorescent protein (GFP) gene. The filler sequence contains stop codons and thus E cells. colique carrying the filler plasmid pCW0051 formed non-fluorescent colonies. The stuffing vector pCW0051 was digested with BsaI and KpnI. A codon library encoding URP sequences of 36 amino acids in length was constructed as shown in Fig. 41. The URP sequence was designated rPEG_136 and had the amino acid sequence (GSGGEG)6. The insert was obtained by hybridizing pairs of synthetic oligonucleotides encoding the amino acid sequence GSGGEGGSGGEG, in addition to a pair of oligonucleotides encoding an adapter with the KpnI site. The following oligonucleotides were used: pr_LCW0057for: AGGTAGTGGWGGWGARGGWGGWTCYGGWGGAGAAGG, pr_LCW0057rev:

[0252]ACCTCCTTCTCCWCCRGAWCCWCCYTCWCCWCCACT, pr_3KpnIstopperFor: AGGTTCGTCTTCACTCGAGGGTAC, pr_3KpnIstopperRev: CCTCGAGTGAAGACGA. The hybridized oligonucleotide pairs were ligated, which produced a mixture of products with variable length representing the variable number of rPEG_J12 repeats. The product corresponding to the length of rPEG_J36 was isolated from the mixture by agarose gel electrophoresis and ligated to the BsaI/KpnI-digested filler vector pCW0051. The majority of clones in the resulting library designated LCW0057 showed green fluorescence after induction, showing that the rPEG_J36 sequence had been ligated in frame with the GFP gene. The iterative screening and multimerization procedure of rPEG_J36 sequences is illustrated in Fig. 14. The present inventors screened 288 strains isolated from the LCW0057 library for high level fluorescence. The 48 isolates with strong fluorescence were analyzed by PCR to verify the length of the rPEG_J segment and 16 clones were identified as having the expected length of rPEG_J36. This procedure produced a collection of 16 rPEG_J36 isolates, which show high expression and differ in their codon usage. The isolated strains were pooled and dimerized using a procedure briefly explained in Fig. 40. A mixture of plasmids was digested with BsaI/NcoI and a fragment comprising the sequence rPEG_J36 and a portion of GFP was isolated. The same plasmid mixture was also digested with BbsI/NcoI and the vector fragment comprising rPEG_J36, the bulk of the plasmid vector, and the remainder of the GFP gene was isolated. Both fragments were mixed, ligated and transformed into BL21 and isolated strains were screened for fluorescence. This dimerization procedure was repeated two more rounds as briefly explained in Fig. 14. During each round, the present inventors doubled the length of the rPEG_J gene and finally obtained a collection of genes encoding rPEG_J288. The amino acid and nucleotide sequence of rPEG_J288 is shown in Fig. 15. It can be seen that the rPEG_J288 module contains segments of rPEG_J36 that differ in their nucleotide sequence despite having identical amino acid sequence. Thus, the present inventors minimized the internal homology in the gene and as a result the present inventors reduced the risk of spontaneous recombination. The present inventors cultivated b L21 of E. colique harbored plasmids encoding rPEG_J288 for at least 20 duplicates and no spontaneous recombination was observed.

Example: Construction of rPEG H288

[0253]A gene library encoding a 288 amino acid URP called rPEG_H288 was constructed using the same procedure that was used to construct rPEG_J288. rPEG_H288 has the amino acid sequence (GSGGEGGSGGSG)24. The flow chart of the construction procedure is shown in Fig. 14. The complete amino acid sequence, as well as the nucleotide sequence of an rPEG_H288 isolate, is given in Fig. 16.

Example: Serum stability of rPEG J288

[0254]A fusion protein containing an N-terminal Flag tag and the rPEG_J288 URP sequence fused to the N-terminus of green fluorescent protein was incubated in 50% mouse serum at 37°C for 3 days. Samples were extracted at various time points and analyzed by SDS-PAGE, followed by detection using Western analysis. An antibody against the N-terminal Flag tag was used for Western detection. The results are shown in Fig. 28, which indicates that a 288 amino acid URP sequence can be completely stable in serum for at least three days.

Example: Absence of preexisting antibodies to rPEG J288 in serum

[0255]The existence of antibodies against URP would be an indication of a possible immunogenic response to this glycine-rich sequence. To test the presence of antibodies existing in serum, a URP-GFP fusion was subjected to an ELISA by immobilizing URP-GFP on a support and subsequently incubating with 30% serum. The presence of antibodies bound to URP-GFP was detected using an anti-IgG-horseradish peroxidase antibody and substrate. The data are shown in Fig. 29. The data show that the fusion protein can be detected by antibodies against GFP or Flag, but not by murine serum. This indicates that the murine serum does not contain antibodies containing the URP sequence.

Example: Purification of a fusion protein containing rPEG J288

[0256]The present inventors purified a protein with the Flag-rPEG_J288-H6-GFP architecture. The protein was expressed in BL21 of E. colien medium SB. Cultures were induced with 0.5 mM IPTG overnight at 18°C. The cells were collected by centrifugation. The pellet was resuspended in TBS buffer containing benzonase and a commercial protease inhibitor mixture. The suspension was heated for 10 min at 75 °C in a water bath to lyse the cells. Insoluble material was removed by centrifugation. The supernatant was purified using immobilized metal ion (IMAC) specificity, followed by a column with immobilized anti-Flag antibody. Fig. 43 shows the PAGE analysis of the purification procedure. The procedure gave protein with at least 90% purity.

Example: Construction of fusion protein between rPEG J288 and interferon-alpha

[0257]A gene encoding human alpha interferon was designed using codon optimization for expression of E. coli.The synthetic gene was fused to a gene encoding rPEG_J288. A His6 tag was provided at the N terminus to facilitate detection and purification of the fusion protein. The amino acid sequence of the fusion protein is given in Fig. 44.

Example: Construction of the rPEG J288-G-CSF Fusion

[0258]A gene encoding human G-CSF was designed using codon optimization for expression of E. coli.The synthetic gene was fused to a gene encoding rPEG_J288. A His6 tag was provided at the N terminus to facilitate detection and purification of the fusion protein. The amino acid sequence of the fusion protein is given in Fig. 44.

Example: Construction of the rPEG J288-hGH Fusion

[0259]A gene encoding human growth hormone was designed using codon optimization for expression of E. coli.The synthetic gene was fused to a gene encoding rPEG_J288. A His6 tag was provided at the N terminus to facilitate detection and purification of the fusion protein. The amino acid sequence of the fusion protein is given in Fig. 44.

Example: Expression of fusion proteins between rPEG J288 and human proteins

[0260]Fusion proteins between rPEG_J288 and two human proteins, interferon-alpha and human growth hormone, were cloned into a T7 expression vector and transformed into BL21 fromE. coli.Cells were grown at 37°C to an optical density of OD 0.5. Subsequently, the cells were cultured at 18 °C for 30 min. Then 0.5 mM IPTG was added and the cultures were incubated in a shaking incubator at 18 °C overnight. Cells were collected by centrifugation and soluble protein was released using BugBuster (Novagen). Both insoluble and soluble protein fractions were separated by SDS-PAGE and the fusion proteins were detected by Western using an antibody against the N-terminal His6 tag for detection. Fig. 45 shows Western analysis of the two fusion proteins, plus rPEG_J288-GFP as a control. All fusion proteins were expressed and the majority of the protein was in the soluble fraction. This is a proof of the high solubility of rPEG_J288 because most attempts at the expression of interferon-alpha and human growth hormone in the cytosol of E. coli, which have been reported in the literature, resulted in the formation of insoluble inclusion bodies. Fig. 45 shows that most of the fusion proteins are expressed as full-length proteins, that is, no fragment was detected suggesting incomplete synthesis or degradation of the partial protein.

Example: Construction and binding of a VEGF multimer

[0261]Libraries of cysteine-restricted peptides were constructed as published [Scholle, M. D. et al. (2005) Comb Chem High Throughput Screen, 8: 545-51]. These libraries were immunopurified against human VEGF and two binding modules consisting of FTCTNHWCPS or FQCTRHWCPI amino acid sequences were identified. Oligonucleotides encoding the amino acid sequence FTCTNHWCPS were ligated to a nucleotide sequence encoding the URP sequence rPEG_A36 with the sequence (GGS)<12>. Subsequently, the fusion sequence was dimerized using restriction enzymes and ligation steps to construct a molecule containing 4 copies of the VEGF binding module separated by rPEG_A36 fused to GFP. The binding affinity for VEGF of fusion proteins containing between zero and four VEGF binding units was compared in Fig. 30. A fusion protein containing only rPEG_A36 fused to GFP shows no affinity for VEGF. The addition of increasing numbers of VEGF-binding modules increases the affinity of the resulting fusion proteins.

Example: Discovery of 1SS binding modules against therapeutic targets

[0262]Random peptide libraries were generated according to Scholle et al. [Scholle, M.D. et al. (2005) Comb Chem High Throughput Screen, 8: 545-51]. Naive peptide libraries displayed cysteine-restricted peptides spaced 4 to 10 random residues apart. The layout of the library is illustrated in the table:

[0263]The libraries were again immunopurified against a series of therapeutically relevant targets using the following protocol: wells on immunosorbent ELISA plates were coated with 5 pg/ml of target antigen in PBS overnight at 4 ° C. The coated plates were washed with PBS, and nonspecific sites were blocked with blocking buffer (PBS containing both 0.5% BSA and 0.5% egg albumin) for 2 h at room temperature. The plates were then washed with PBST (PBS containing 0.05% Tween 20), and phage particles at 1-5x1012/ml in binding buffer (blocking buffer containing 0.05% Tween 20) were added. to the wells and incubated with shaking for 2 h at room temperature. The wells were then emptied and washed with PBST. Bound phage particles were eluted from the wells by incubation with 100 mM HCl for 10 min at room temperature, transferred to sterile tubes, and neutralized with 1 M TRIS base. For infection, SS320 ofE. Log phase colides grown in Super Broth supplemented with 5 pg/ml tetracycline were added to the neutralized phage eluate, and the culture was incubated with shaking for 30 min at 37 °C. Infected cultures were then transferred to larger tubes containing Super Broth with 5 pg/ml tetracycline and the cultures were incubated with shaking overnight at 37°C. Overnight cultures were cleared of E. coli by centrifugation, and the phage was precipitated in the supernatant after the addition of a solution of 20% PEG and 2.5 M NaCl to a final PEG concentration of 4%. The precipitated phage was collected by centrifugation, and the phage pellet was resuspended in 1 ml of PBS, cleaned of E. coliresidual by centrifugation and transferred to a new tube. Phage concentrations were estimated spectrophotometrically and the phage was used for the next round of selection. Individual clones were screened for target binding affinity after 3 or 4 rounds of phage immunopurification. Individual plaques of phage clones selected during immunopurification were collected in Super Broth containing 5 pg/ml tetracycline and grown overnight with shaking at 37°C. ELISA plates were prepared by coating antigen and control proteins (BSA, egg albumin, IgG) at 3 pg/ml in PBS overnight at 4 °C. The plates were washed with PBS and blocked with blocking buffer (PBS containing 0.5% BSA) for 2 h at room temperature. Overnight cultures were cleared of E. colipor centrifugation and the supernatant was diluted 1:10 in binding buffer (blocking buffer containing 0.05% Tween 20) and transferred to ELISA plates after washing with PBST (PBS containing 0.05% Tween twenty). The plates were incubated with shaking for 2 h at room temperature. After washing with PBST, anti-M13-HRP (Pharmacia), 1:5000 dilution in PBS, was added to wells. The plates were incubated with shaking for 30 min at room temperature and washed with PBST, followed by PBS. A substrate solution containing 0.4 mg/ml ABTS and 0.001% H<2>O<2>in 50 mM phosphate-citrate buffer was added to the wells, and allowed to develop for 40 min after The plates were read in a plate reader at 405 nm. These ELISA reads allowed determination of clone specificity, and antigen-specific clones were commercially sequenced using established procedures.

Table X: Specific binding module sequences for EpCAM

Table X: Sequences of specific binding modules for VEGF

W E C T Q H W C P S sNG0025S3.021 A P F F S C S F G F C R D L Q T sNG0026S3.035 T P Y F R C Q F G F C F D S F S sNG0026S3.045 N P F F Y C V A G K C V D A P L sNG0026S3.029

Table 0028S3.086 S L S Q Q S C D W C W L F G P P sNG0028S3.060 K R L L E C G A L C A L F G P P sNG0028S3.008 H T I L T C D S G F C T L F G P sNG0028S3.012 N L W H V C H T S L C H S R L A sNG0028S3.0 92 N S F Y L C H S S V C G Q L P S sNG0028S3.082 A G F S C E N Y F F C P P K N L sNG0028S3. 016 S W C T V F G N H D P S C N S R sNG0028S3.004

C S S N G R W K A H C sNG0028S3.076 L P N M W R V V V P D V Y D R R sNG0028S3.068

Table

Table

Table

Table 034S3.087 Y Y L T N C P K G H C L R R V D sNG0034S3.080

(continuation)

Table X: Specific binding module sequences for DR5

L Y L H S C S R G I C L S P R V sNG0034S3.082 F S C Q S S F P G R R M C E L R sNG0034S3.040 H R C S A H G S S S S S F C P G S sNG0034S3.029

Table X: Sequences of specific binding modules for TrkA

K T W D C N S G H C V J T F K sNG0035S3.074 A T W D C R D H N F S C V R L S sNG003SS3.089

Example: aEpCAM drug conjugates

[0264]Anti-EpCAM peptides were isolated from random peptide libraries that were generated according to Scholle et al. [Scholle, M.D. et al. (2005) Comb Chem High Throughput Screen, 8: 545-51]. Naive peptide libraries visualized cysteine-restricted peptides with cysteines separated by 4 to 10 random residues. After three rounds of affinity selection with the above libraries several peptide ligands specific for EpCAM (EpCam1) were isolated (Table X). EpCam1 isolates have a conserved cysteine gap of four amino acids (CXXXXC). The EpCam1 peptide ligands were then gently randomized (except cysteine positions) with codons encoding 3-9 residues and moved into a phagemid vector. The phagemid libraries were subsequently affinity-screened against EpCAM to isolate peptide ligands optimized for binding (Table X, EpCam2). EpCam2 ligands contain the conserved CXXXXC cysteine cleavage. Furthermore, most anti-EpCam sequences do not contain a lysine residue, which allows conjugation with free amine groups outside the binding sequences. Additionally, anti-EpCam peptide ligands can be genetically fused to URP sequences (of any length) and multimerized using iterative dimerization. The resulting anti-EpCAM MURPs can be used to specifically target EpCAM with high affinity for monomer sequences. An example of a tetramer EpCAM-URP amino acid sequence is shown in Fig. 31. This sequence only contains two lysine residues that are located at the N-terminal Flag tag. The side chains of these lysine residues are particularly suitable for drug conjugation.

Table X. Anti-EpCam sequences

Name Sequence

EpCam 1 LRCWGMLCYA

LRCIGQICWR

LKCLYNICWV

LFCWGNVCHF

LTCWGQVCFR

RPGMACSGQLCWLNSP

- PHALQCYGSLCWPSHL

RAGITCHGHLCWPITD

RPALKCIGTLCSLANP

PHGLWCHGSLCHYPLA

PHGLICAGSICFWPPP

PRNLTCYGQICFQSQH

PHNLACQNSICVRLPR

PHGLTCTNQICFYGNT

EpCam 2 HSLTCYGQICWVSNI

PTLTCYNQVCWVNRT

PALRCLGQLCWVTPT

PGLRCLGTLCWVPNR

RNLTCWNTVCYAYPN

RGLKCLGQLCWVSSN

PTLKCSGQICWVPPP

RNLECLGNVCSLLNQ

PTLTCLNNLCWVPPQ

RGLKCSGHLCWVTPQ

HGLTCHNTVCWVHHP

HTLECLGNICWVINQ

HGLTCYNQICWAPRP

HGLACYNQLCWVNPH

(continuation)

Name Sequence

RGLACQGNICWRLNP

RAITCLGTLCWPTSP

LTLECIGNICYVPHH

Example: Adding random sequences

[0265]The binding modules can be affinity matured, or elongated, by the addition of URP-like linkers and random sequence to the N terminus, C terminus, or both the N and C termini, of the binding sequence. Fig. 32 shows the addition of naïve cysteine-restricted sequences to an anti-EpCAM binding module. Libraries of random sequence additions can be generated using a single-stranded or double-stranded DNA cloning approach. Once generated, libraries can be selected by affinity against the initial target protein or a second protein. For example, an addition library containing an anti-EpCAM binding module can be used to select sequences containing 2 or more binding sites to the target protein.

Example: Building a 2SS Training Library

[0266]A series of oligonucleotides were designed to construct a library based on the VEGF binding peptide 1SS FTCTNHWCPS. The oligonucleotides incorporated variations in the cysteine distance patterns of the flanking sequences, while the VEGF-binding peptide sequence was kept fixed.

Direct oligonucleotides:

[0267]LMS70-1 CAGGCAGCGGGCCCGTCTGGCCCGTGYTTTACTT GTACGAATCATT GGT GTCCT

[0268]LMS70-2 CAGGCAGCGGGCCCGTCTGGCCCGTGYN NKTTTAACTTGTACGAATCATTGGTGTCCT[0269]LMS70-3

CAGGCAGCGGGCCCGTCTGGCCCGTGYNNKNNKTTT ACTT GT ACGAATCATT GGT GTCCT

[0270]LMS70-4 CAGGCAGCGGGCCCGTCTGGCCCGTGYNHTNHTTTTACTTGTACGAATCATTGGTGTCCT

[0271]LMS70-5

CAGGCAGCGGGCCCGTCTGGCCCGTGYN HTNHTNHTNH'mTACTTGTACGAATCATTGCrrG TCCT

[0272]LMS70-6

CAGGCAGCGGGCCCGTCTGGCCCGTGYKMTKMTKMTKMTKMTTTTACTTGTACGAATCAT

Reverse Oliqonucleotides (reverse complemented):

[0273]LMS70-1R ACCGGAACCACCAGACTGGCCRCACGAAGGACACCAATGATTCGT ACAA

[0274]LMS70-2R ACCGGAACCACCAGACTGGCCRCAMNNCGAAGGACACCAATGATTCGTACAA[0275]LMS70-3R ACCGGAACCACCAGACTGGCCRCAMNNMNNCGAAGGACACCAATGATTCGTACAA

[0276]LMS70-4R ACCGGAACCACCAGACTGGCCRCAAADNADNDNCGAAGGACACCAATGATTCGTACAA

[0277]LMS70-5R ACCGGAACCACCAGACTGGCCRCAAADNADNADNDNCGAAGGACACCAATGATTCGTACAA

[0278]LMS70-6R

ACCGGAACCACCAGACTGGCCRCAAKMAKMAKMAKMAKMCGAAGGACACCAATGATTCG

TACAA

Oligonucleotide dilutions

[0279]Mixture 1 (from 100 pM stock mixes): 100 pl of 70-6, 33 pl of 70-5, 11 pl of 70-4, 3.66 pl of 70-3, 1.2 pl of 70- 2, 0.4 pl of 70-1. Mixture 2 (from 100 pM stock mixes): 100 pl of 70-6R, 33 pl of 70-5R, 11 pl of 70-4R, 3.66 pl of 70-3R, 1.2 pl of 70-2R, 0 .4 pl of 70-1R

PCR assembly

[0280]10.0 µl template oligonucleotide (5 pM), 10.0 µl 10X buffer, 2.0 dNTPs (10 mM), 1.0 µl cDNA polymerase (Clonetech), 77 µl H<2 >Or distilled. PCR program: 95 °C 1 min, (95 °C 15 s, 54 °C 30 s, 68 °C 15 s) x5, 68 °C 1 min

PCR amplification

[0281]Primers, 10.0 µl of assembled mixture, 10.0 µl of 10 pM), 1.0 pl of cDNA polymerase (Clonetech), 57 pl of distilled H<2>O. PCR program: 95 °C 1 min, (95 °C 15 s, 54 °C 30 s, 68 °C 15 s) x25, 68 °C 1 min. The product was purified by Amicon Y10 column. The assembled product was digested with Sfil and BstXI and ligated into the phagemid vector pMP003. Ligation was performed overnight at 16°C in an MJ PCR machine. The ligation was then purified by precipitation with EtOH. Transformation into fresh competent ER2738 cells by electroporation.

[0282]The resulting library was immunopurified against VEGF as described below. Several isolates were identified that showed improved VEGF binding over the 1SS starting sequence. Binding and expression data are shown in Fig. 38. Sequences and results of Western analysis of training clones are shown in Fig. 39.

Example: Phage immunopurification of training libraries

[0283]First round of immunopurification:

[0284]1) First round, coat 4 wells per library to be screened. Coat the well of a Costar 96-well ELISA plate with 0.25 pg of VEGF<121>antigen in 25 μl of PBS. Cover the plate with a plate sealant. Coating can be done overnight at 4°C or for 1 h at 37°C.

[0285]2) After shaking off the coating solution, block the well by adding 150 µl of PBS/1% bSa. Seal and incubate for 1 h at 37°C.

[0286]3) After shaking off the blocking solution, add 50 µl of freshly prepared phage (see library reamplification protocol) to the well. Just for the first round, also add 5 pl of 5% Tween. Seal the plate and incubate for 2 h at 37°C.

[0287]Meanwhile, inoculate 2 ml of SB medium plus 2 µl of 5 mg/ml tetracycline with 2 µl of an ER 2738 cell preparation and allow growth at 250 rpm and 37 °C for 2.5 h. Grow 1 culture for each library being screened that includes negative selections. Take all precautions to avoid contamination of the culture with phage.

[0288]4) Shake the phage solution, add 150 µl of PBS/0.5% Tween to the well and pipet 5 times vigorously up and down. Wait 5 min, shake and repeat this washing step. In the first round, wash this way 5 times, in the second round 10 times, and in the third, fourth and fifth rounds 15 times.

[0289]5) After shaking off the final wash solution, add 50 µl of 10 mg/ml trypsin in freshly prepared PBS, seal and incubate for 30 min at 37 °C. Pipet 10 times vigorously up and down and transfer the eluate (4 x 50 pl in the first round, 2 x 50 ml in the second round, 1 x 50 pl in subsequent rounds) to the 2 ml culture of E. coliprepared and incubate at room temperature for 15 min.

[0290]6) Add 6 ml of prewarmed SB medium, 1.6 pl of carbenicillin and 6 pl of 5 mg/ml tetracycline. Transfer the culture to a 50 ml polypropylene tube.

[0291]7) Shake the 8 ml culture at 250 rpm and 37 °C for 1 h, add 2.4 μl of 100 mg/ml carbenicillin and shake for an additional hour at 250 rpm and 37 °C.

[0292]8) Add 1 ml of VCSM 13 helper phage and transfer to a 500 ml polypropylene centrifuge bottle. Add 91 ml of prewarmed SB medium (37 °C) and 46 μl of 100 mg/ml carbenicillin and 92 μl of 5 mg/ml tetracycline. Shake the 100 ml culture at 300 rpm and 37 °C for 11/2 to 2 h.

[0293]9) Add 140 μl of 50 mg/ml kanamycin and continue stirring at 300 rpm and 37 °C overnight.

[0294]10) Centrifuge at 4000 rpm for 15 min at 4 °C. Transfer the supernatant to a clean 500 ml centrifuge bottle and add 25 ml of 20% PEG-8000/2.5 M NaCl. Store on ice for 30 min.

[0295]11) Centrifuge at 9000 rpm for 15 min at 4 °C. Discard the supernatant, drain inverted onto a paper towel for at least 10 min, and wipe the remaining liquid from the top of the centrifuge bottle with a paper towel.

[0296]12) Resuspend the phage pellet in 2 ml of PBS/0.5% BSA/0.5% Tween buffer by pipetting up and down along the side of the centrifuge bottle and transfer to a tube. 2 ml microcentrifuge. Resuspend further by pipetting up and down using a 1 mL pipette tip, centrifuge at full speed in a microcentrifuge for 1 min at 4 °C, and transfer the supernatant to a 0.2 μm filter in a sterile 2 mL microcentrifuge tube. .

[0297] 13) Continue step 3) during the next round or store the phage preparation at 4 °C. 0.02% (w/v) sodium azide may be added for long-term storage. Only freshly prepared phage should be used for each round.

[0298] Second round of immunopurification

[0299] Second round, coat 2 wells per library to be screened. Coat the well of a Costar 96-well ELISA plate with 0.25 pg of VEGF<121>antigen in 25 μl of PBS. Cover the plate with a plate sealant. Coating can be done overnight at 4°C or for 1 h at 37°C.

[0300] Also block 2 uncoated wells for each library that is to be used as a negative control for the enrichment ratio calculation.

[0301] Third round of immunopurification

[0302] Third round, coat 1 well per library to be screened. Coat the well of a Costar 96-well ELISA plate with 0.25 pg of VEGF121 antigen in 25 µl of PBS. Cover the plate with a plate sealant. Coating can be done overnight at 4°C or for 1 h at 37°C.

[0303] Also block 1 uncoated wells for each library that is to be used as a negative control for the enrichment ratio calculation.

Example: Solution-based immunopurification:

[0304] 1. Biotinylate the target protein according to the manufacturer.

[0305] 2. Coat a total of 8 wells (per selection) with 1.0 pg neutravidin (Pierce) in PBS and incubate overnight at 4 °C.

[0306] 3. Block the wells with SuperBlock (Pierce) for 1 h at room temperature. Store the plate with blocking buffer until needed (in step 6).

[0307] 4. Use 100 nM of biotinylated target protein and add 1012 phage/ml (in PBST) for a total volume of 100-200 µl using SuperBlock plus 0.05% Tween 20.

[0308] 5. Turn the phage-target mixture at room temperature for at least 1 h.

[0309] 6. Dilute 100 µl of phage-target mixture with 700 µl of SuperBlock, mix and add 100 µl to each of 8 neutravidin-coated wells (from step 3).

[0310] 7. Incubate for 5 min at room temperature.

[0311] 8. Wash 8X with PBST.

[0312] 9. Elute the phage with 100 μl of 100 mM HCl for 10 min.

[0313] 10. Neutralize by adding 10 μl of 1 M TRIS at pH=8.0.

[0314] 11. Infect the cells for seeding or amplify the phage for a subsequent round of solution immunopurification.

Example: Phage ELISA screening for VEGF-positive clones

[0315] 1) Add 0.5 ml of SB containing 50 pg/ml carbenicillin to deep 96-well plate. Collect a colony and inoculate wells.

[0316]2) Shake the plate containing the bacterial cultures at 300 rpm overnight at 37 °C.

[0317]3) Prepare 4 ng/pl of target protein solution in PBS. Add 25 µl (100 ng) of protein to each well and incubate overnight at 4 °C.

[0318]4) Shake the coated ELISA plates and wash 2x with PBS. Add 150 µl/well of PBS 0.5% BSA (blocking buffer). Block for 1 h at RT.

[0319]5) Centrifuge the microtube racks (3000 rpm; 20 min).

[0320]6) Prepare binding buffer (0.5% Tween 20 blocking buffer). Take 135 µl aliquots of binding buffer per well into a low protein binding 96-well plate.

[0321]7) Shake the wells onto ELISA plates and wash 2 times with PBST (PBS 0.5% Tween 20).

[0322]8) Dilute 15 µl of phage from overnight cultures 1:10 in PBST, mix by pipetting and transfer 30 µl to each protein-coated well. Incubate for 2 h at RT with gentle shaking.

[0323]9) Wash the plates 6 times with PBST.

[0324]10) Add 50 µl of anti-M13-HRP 1:5000 in binding buffer to the wells. Incubate for 30 min with gentle shaking at RT.

[0325]11) Wash the plates 4 times with PBST, followed by 2 times with H<2>O.

[0326]12) Prepare 6 ml of ABTS solution (5.88 ml of citrate buffer plus 120 pl of ABTS and 2 pl of H<2>O<2>). Take 50 µl aliquots per well on each ELISA plate.

[0327]13) Incubate at RT and read the OD at 405 nm using an ELISA plate reader at appropriate time points depending on the signal (up to 1 h)

Example: Dimming of junction modules

[0328]Phage-displayed libraries of 10e9 to 10e11 cyclic peptides were created with 4, 5, 6, 7, 8, 9, 10, 11 and 12 amino acids randomized or partially randomized between disulfide-linked cystines, and in some cases amino acids additional randomizations on the exterior of the cysteine pair by conventional procedures. Immunopurification of these cyclic peptide libraries against several targets, including human VEGF, reliably yielded peptides that specifically bound hVEGF and not BSA, egg albumin, or IgG.

Example: Construction and immunopurification of a plexin-based library

[0329]Two libraries were designed based on the plexin scaffold. The Pfam protein database was used for phylogenetic alignment of naturally occurring plexin domains as shown in Fig. 35. The central part of the plexin scaffold (Cys24-Gly25-Trp26-Cys27) is conserved in both designs. library and served as a crossover region for the generation of N and C libraries. The randomization schemes of both plexin libraries are shown in Fig. 36. The two libraries were generated by overlapping two oligonucleotides encoding libraries in the crossover region and using pull-through PCR, followed by restriction cloning (Sfil/BstXI) and cloning into the phagemid vector pMP003. The resulting plexin libraries were designated LMP031 (N-terminal library) and LMP032 (C-terminal library) and each was represented by a complexity of approximately 5 × 108 independent transformants. For validation, approximately 24 Carb-resistant clones from each unselected library were analyzed by PCR. Clones that gave a fragment of the correct size (375 bp) were further analyzed by DNA sequencing. Correct full-length plexin sequences were obtained for 50% and 67% of clones derived from the LMP031 and LMP032 libraries, respectively.

[0330]The two libraries were mixed together at a 50/50 ratio and immunopurified in parallel against VEGF, death receptor Dr4, ErbB2 and HGFR immobilized on 96-well ELISA plates. Four rounds of immunopurification were carried out using 1000 ng of target protein in the first round, 500 ng in the second round, 250 ng in the third round, and 100 ng in the fourth round. After the final round of immunopurification, 192 Carb-resistant clones from each screen were tested for binding to 100 ng of immobilized target protein, human IgG, egg albumin, and BSA by phage ELISA using peroxidase-conjugated polyclonal anti-M13 Ab. horseradish for detection. The highest percentage of positive clones was obtained for target DR4 (69%), followed by target ErbB2 (53%), HGFR (13%) and BoNT target (1%). Positive clones were further analyzed by PCR and DNA sequencing. All clones revealed unique sequences and all but one (against DR4) were derived from LMP032 (C-terminal library). The sequences of some of the identified target-selective isolates are shown in Fig. 37.

[0331]For further analysis, a selection of selected target-specific binders were first subcloned into the protein expression vector pVS001, then produced as soluble microproteins, and finally purified by thermal lysis. Purified target-specific microproteins are analyzed by protein ELISA to confirm target recognition, by SDS-PAGE to confirm monomer formation, and by surface plasmon resonance to measure their affinities for the target. The best clones are used in the next round of library generation to further improve their properties.

Example: Building a library based on snake venom

[0332]Phage-displayed libraries of 10e8 to 10e10 of the 3-finger toxin (3FT) scaffold were created with partially randomized amino acids from the tip of finger 1 and the descending part of finger 2 or the tip of finger 3 and the ascending finger 2 using conventional procedures.

[0333]Two 3FT scaffolds were used as templates for the generation of the 3FT library (finger configuration 1 and 2). The structure of a 3FT scaffold and a multiple sequence alignment of related sequences is shown in Fig. 33. A library was designed such that two surface loops of the toxin are randomized as illustrated in Fig. 34. The A partially randomized 3FT scaffold library was generated by overlapping four oligonucleotides encoding the library in the hybridization regions and using stretch-through PCR (puM-through), followed by restriction cloning (Sfil/BstXI) into the vector. phagemid pMP003. The resulting 3FT library was designated LMP041.

Example: Grafting binding peptides onto microprotein scaffolds – target-specific peptide-assisted randomization

[0334]The goal here is to use peptides that have been identified to be specific for targets of interest in order to generate 3SS plus target-specific binders. This strategy is illustrated using VEGF-specific peptide transfer into the tip of finger 1 of the 3FT scaffold and modifying the AA residues of finger 2, which are in close proximity to the target-specific sequence to generate high-affinity VEGF binders. . Phage display libraries 10e8 to 10e10 of the 3-finger toxin (3FT) scaffold with VEGF-specific sequence of finger tip 1 and partially randomized downstream part of finger 2 were created by standard procedures as described in the example above, except that 2 random finger 1 forward primers were replaced with F1-VEGF-specific forward primer encoding the following sequence: P S G P S C H T T N H W P I S A V T C P P.

[0335]The 3FT scaffold library (specific for VEGF) centered with partially randomized finger 2 was generated by overlapping four oligonucleotides encoding the library in the hybridization regions and using pull-through PCR, followed by restriction cloning (SfiI/BstXI) into the phagemid vector pMP003. The resulting 3FT library was designated LMP042.

Example: Plasma half-life of a MURP

[0336]The plasma half-life of MURP can be measured after i.v. injection. or i.p. of MURP in catheterized rats essentially as described by [Pepinsky, R. B. et al. (2001) J Pharmacol Exp Ther, 297: 1059 66]. Blood samples can be drawn at various time points (5 min, 15 min, 30 min, 1 h, 3 h, 5 h, 1 d, 2 d, 3 d) and the plasma concentration of MURP can be measured using ELISA . Pharmacokinetic parameters can be calculated using WinNonlin version 2.0 (Scientific Consulting Inc., Apex, NC). To analyze the effect of the URP module, the plasma half-life of a protein containing the URP module can be compared with the plasma half-life of the same protein lacking the URP module.

Example: Solubility test of a MURP

[0337] The solubility of MURP can be determined by concentrating purified samples of MURP in physiological buffers such as phosphate-buffered saline at various concentrations in the range of 0.01 mg/ml to 10 mg/ml. Samples can be incubated for up to several weeks. Samples where the concentration exceeds the solubility of MURP show precipitation as indicated by turbidity, which can be measured in an absorbance reader. Precipitated material can be removed by centrifugation or filtration and the remaining protein concentration in the supernatant measured using a protein assay such as the Bradford assay by measuring absorbance at 280 nm. Solubility studies can be accelerated by freezing samples at -20°C and subsequent thawing. This procedure frequently leads to the precipitation of poorly soluble proteins.

Example: Serum binding activity of MURP

[0338] MURPs of interest can be coated on microtiter plates and control proteins on other wells of the plate. Subsequently, serum samples of interest can be added to the wells for 1 hour. The wells can then be washed with a plate washer. Bound serum proteins can be detected by adding antibodies against serum proteins that have been conjugated with enzymes such as horseradish peroxidase or alkaline phosphatase for detection. Another way to detect serum binding to MURP, add the MURP of interest to serum for approximately 1 hour to allow binding. MURP can then be immunoprecipitated using an antibody against an epitope in the MURP sequence. Precipitated samples can be analyzed by PAGE and optionally by Western to detect any proteins that co-precipitated with MURP. Serum proteins showing co-precipitation can be identified by mass spectrometry.

Claims (10)

1. Procedure for increasing the serum half-life of a protein, comprising: fusing said protein with one or more unstructured recombinant polymers (URP), wherein the URP comprises at least about 200 contiguous amino acids, and wherein (a) The sum of the glycine (G), aspartate (D), alanine (A), serine (S), threonine (T), glutamate (E) and proline (P) residues contained in the URP constitutes more than approximately 80% of the total amino acids of the URP; and (b) at least 50% of the amino acids of the URP are not present in the secondary structure as determined by the Chou-Fasman algorithm; and wherein said URP is substantially incapable of binding non-specifically to a whey protein; and (c) the URP has a T epitope score equal to or less than -4. 2. Method of claim 1, wherein the serum half-life of the protein is prolonged by at least 2-fold. 3. Method of claim 1, wherein the URP comprises a non-natural amino acid sequence. 4. Method of claim 1, wherein the URP substantially lacks secondary structure as determined by the Chou-Fasman algorithm. 5. Method of claim 1, wherein glycine residues contained in the URP constitute at least about 50% of the total amino acids of the URP. 6. Method of claim 1, wherein the URP comprises sequence repetition. 7. Method of claim 1, wherein the protein is a pharmaceutically active protein. 8. Method of claim 1, wherein the protein comprises one or more modules selected from the group consisting of binding modules, effector modules, multimerization modules, C-terminal modules and N-terminal modules. 9. Method of claim 7, wherein the pharmaceutically active protein is selected from the group consisting of cytokines, growth factors, enzymes, receptors, microproteins, hormones, erythropoietin, adenosine deiminase, asparaginase, arginase, interferon, growth hormone , growth hormone releasing hormone, G-CSF, GM-CSM, insulin, hirudin, TNF receptor, uricase, rasburicase, axokine, RNase, DNase, phosphatase, Pseudomonas exotoxin, ricin, gelonin, desmoteplase, laronidase, thrombin, enzyme blood coagulation, VEGF, protropin, somatropin, alteplase, interleukin, factor VII, factor VIII, factor 10. Method of claim 1, wherein any one type of amino acids alone selected from the group consisting of glycine (G), aspartate (D), alanine (A), serine (S), threonine (T), glutamate (E) and proline (P) contained in the URP constitute more than about 20% or more of the total amino acids of the URP.